·No text on this page.

·No text on this page.

·No text on this page.

·ROYAL SOCIETY,
REPORT
TIIE COMMITTEE OF
PHYSICS, INCLUDING METEOROLOGY,
ON THE
OBJECTS OF SCIENTIFIC INQUIRY
IN THOSE SCIENCES.
APPROVED BY
TITE PRESIDENT AND COUNCIL.
LONDON:
PRINTED BY RICHARD AND JOHN & TAYLOR,
RKD LION COURT, FLERT ETRHET.
LION
1840.

·No text on this page.

·CONTENTS. ...................................................... re ....................................................... ........................................................................ PHYSICS AND METEOROLOGY. Page 1. Terrestrial Magnetism 2. Figure of the Earth 38 3. Tides 39 4. Meteorology 5. Distribution of Temperature in the Sea and Land....... 48 6. Currents of the Ocean ........................................................... 49 7. Depth of the Sea 51 8. Variable Stars 51 9. Refraction 52 10. Eclipses 52 ................. ......... .................... ...... ................. ..................... ..................... INSTRUCTIONS FOR MAKING METEOROLOGICAL OBSERVATIONS. 1. Barometers....... 2. Thermometers 3. Actinometers 4. Radiating Thermometers 5. Hygrometers 6. Vancs, Anemometers, and Rain-gauges 7. Clouds and Meteors 8. Electrometers... 9. Registers..... 53 58 61 69 69 71 73 74 76 ............................................ ................................ ......... ............... ................... .................. .................... APPENDIX. Table I.–Correction to be added to Barometers for Capillary Action 81 Table II.--Reduction of Thermometer to 32° Fahrenheit 82 Table III.-Reduction for Barometers 88 Table IV.-Force of Wind.... 88 Table V.-Force of Aqueous Vapour 89 Letters from the Baron Alexander von flumboldt to the Earl of Minto 91 Letter from Professor A. Erman to Major Sabine, R.A., F.R.S. 102 Coinparisons made at the Royal Society's Apartments, Somerset-IIouse, on the 16th, 17th, 19th and 20th of August, 1839, with the Royal Society's Standard, of the Marine, Standard, and Mountain Barometers supplied to the Antarctic Expedition, and to the four fixed Magnetic Observatories of Canada, Cape of Good-Hope, St. Helena, and Van Dicmen's Land: by J. D. Roberton, Assist. Sec. Royal Society 110 Comparisons of Mountain Barometers corresponding with the above Standards.. 110 Comparisons made at the Royal Society's Apartments, Somerset-House, in August 1839 11) Comparisons made in March 1840 111 Comparisons of Standards with the Royal Society's, made for the Honourable the East India Company, in April 1840 111 Comparisons made with the Royal Society's Standard, of various Standard and Mountain Barometers, for the Honourable the East India Company, made in February 1840.... 112 Comparisons of Mountain Barometers corresponding with the above Standards ................................................................ 112 Memorandum on the Books to be kept at the Magnetic Observatories. .... ......... ...... ...... ........ .............. ........ ...........

·No text on this page.

·W! 187) REPORT. PHYSICS AND METEOROLOGY. Tune Committee of Physics, including Meteorology, are very strongly impressed with the number and importance of the desiderata in playsical and meteorological science, which may wholly or in part be supplied by observations which may be expected from the zeal and industry of travellers, residents in foreign countries, and others whose position may give them favourable opportunities of stimulating research on the part of those in their respective departments. While they wish therefore to omit nothing in their enumeration of those objects which appear to them descrving of attentive inquiry on sound scientific grounds, and from which consequences may be drawn of real importance, either for the settlement of disputed questions, or for the advancement of knowledge in any of its branches,—they deem. it equally their duty to ornit or pass lightly over several points which, although not without a certain degree of interest, mny yet be rcgarded in the present state of science rather as matters of abstract curiosity that as affording data for strict reasoning; as well as others, which may be equally well or better elucidated by inquiries instituted at home and at lcisure. 1. TERRESTRIAL MAGNETISM. Terrestrial Magnetism will be considered : 1st, as regards those accessions to our knowledge which may be supplied by observations to be made, independently of any concert with or co-operation of other observers; and 2ndly, as regards those which depend on and require such concert; and are therefore to bc considered with reference to the observations about to be carricd on simultaneously in the fixed magnetic observatories established by Her Majesty's Government, and in the other similar observatories, both public and private, in Europe, India, and elsewhere, with which it is intended to open and maintain a correspondence. Now it may be observed, that these two classes of observations naturally refer themselves to two chief branches into which the B ·2 science of terrestrial magnetism in its present state may be subdivided, and which bear a certain analogy to the thcories of the elliptic movements of the planets, and of their periodical and secular perturbations. The first comprehends the åctual distribution of the magnetic influence over the globe, at the present epoch, in its mean or average state, when the effects of temporary fluctuations are either neglected, er eliminated by extending the observations over a sufficient time to neutralize their effects. The other comprises the history of all that is not permanent in the phenomena, whether it appear in the form of momentary, daily, monthly, or annual change and restoration, or in progressive changes not compensated by counter changes, but going on continually accumulating in one direction, so as in the course of many years to alter the mean amount of the quantities observed. These last-mentioned changes hold the same place, in the analogy above alluded to, with respect to the mean quantities and temporary fluctuations, that the secular variations in the planetary movements must be regarded as holding, with respect to their mean orbits on the one hand, and their perturbations of brief period on the other. There is, huwever, this difference, that in the planetary theory all thcsc varicties of effect have been satisfactorily traced up to a single cause, whereas in that of terrestrial magnetisni this is so far from being demonstrably the case, that the contrary is not destitute of considerable probability. In fact, the great features of the magnetic curves, and their general displacements and changes of form over the whole surface of the carth, would seem to be the result of causes acting in the interior of the earth, and pervading its wholc mass; while the annual and diurnal variations of the needlc, with their train of subordinate periodical movements, may, and very probably do, arise from and correspond to electric currents produced by periodical variations of temperature at its surface, due to the sun's position above the horizon, or in the ecliptic, modificd by local causes ; while local or temporary electric discharges, due to thermic, chemical, or mechanical causes, acting in the higher regions of the atmosphere, and relieving themselves irregularly or at intervals, may serve to render account of those unceasing, and as they scem to us casual movements, which recent observations have placed in so conspicuous and interesting a light. The electrodynamic theory, which refers all magnetism to electric currents, is silent as to the causes of those currents, which may be various, and which only the analysis of their effects can teach us to regard as internal, superficial, or atmospheric It is not merely for the use of the navigator that charts, giving a general view of the lines of Magnetic Declination, Inclination, and Intensity, are necessary. Such charts, could they really be depended on, and were they in any degree complete, would be of the most eminent use to the theoretical inquirer, not only as general directions in the choice of empirical formulæ, but as powerful instruments for facilitating numerical investigation, by the choice ·they afford of data favourably arranged; and above all, as affording decidedly the best means of comparing arry given theory with oba servation. In fact, upon the whole, the readiest, and beyond coparison the fairest and most effectual mode of testing the numerical applicability of a theory of terrestrial magnetism, would be, not servilely to calculate its results for given localities, however numerous, and thereby load its apparent errors with the real errors, both of observation and of local magnetism; but to compare the totality of the lines in our charts with the corresponding lincs, as they result from the formulæ to be tested, when their general agreement or disagrecment will not only show how far the latter truly represent the facts, but will furnish distinct indications of the modifications they rcquire. Unfortunately for the progress of our theories, however, we are yet very far from possessing charts even of that one element, the Declination, most useful to the navigator, which satisfy these requisites; while as respects the others (the Inclination and Intensity) the most lamentable deficiencies occur, especially in the Antarctic regions. To make good thcsc deficiencies by the continual practice of cvery mode of observation appropriate to the circumstances in which the observer is placed throughout a voyage, will be one of the great objects to which attention must bc directed. And first At sca.--We are not to expect from magnetic observations made at sea the precision of which they are susceptible on land. Neverthieless, it has been ascertained that not only the Declination, but the Inclination and Intensity can be observed, in moderate circumstances of weather and sea, with sufficient correctness, to afford most useful and valuable information, if paticnce be bestowed, and proper precautions adopted. The total intensity, it is ascer: tained, can be measured with some considerable degrec of certainty by the adoption of a statical method of observation recently devised by Mr. Fox. And when it is recollected that but for such observations the whole of that portion of the globe which is covered by the occan must remain for ever a blank in our charts, it will be ncedless further to insist on the necessity of making a daily series of magnetic obscrvations, in all the three particulars abovemo oned, whenever weather and sea will permit, an essential re in the business of a voyage. Magnetic observations at sca will, of course, be affected by the ship's magnetism, and this must be eliminated to obtain results of any service. To this end, First. Every scrics of observations made on board should be accompanied with a notice of the direction by compass of the ship's head at the time. Secondly. Previous to sailing, a very careful series of the apparent deviations, as shown by two compasses permanently fixed, (the one as usual, the other in a convenient position, considerably more forward in the ship,) in every position of the ship's head, as compared with the real position of the ship, should be made and recorded, with a view to attempt procuring the constants of the ship's action ac. B 2 ·4 cording to M. Poisson's theory*; and this process should be repeated on one or more convenient'occasions during thevoyage; and, generally, while at anchor, every opportunity should be taken of swinging round the ship's head to the four cardinal points, and executing in each position a complete series of the usual observations. Thirdly. Wherever magnetic instruments are landed and observations made on terra firma, or on ice, the opportunity should be * Note on M. Poisson's theory of the deviation produced in the direction of the compass by the iron of the ship. с -A (Ship's Stern).....G * (Ship's Hcad.) N (North.) Let G x bc thc axis of the ship, æ lying towards the ship's head. G the centre of gravity of the coinpass. G N the meridian, N the north. Let A be the south pole of the dipping needle, that is the extremity of the needle which dips beneath the horizontal plane in onr hemisphere, let C be the projection of the point A upon a horizontal plane, so that G C is the magnetic meridian. Let the angle A G C = 0, 0 varying from - 90° to + 90°, being positive when the south pole A is beneath the horizontal plane, and negative in the contrary casc. Let CGN=4, this angle (variation) which is the azimuth of the vertical plane CG A may extend from - 180° to + 180°, and is to be considered positive or negative according as the live GC falls to the west or to the cast of the linc G N, or this angle may be considered to vary from 0 to 360°, going from north to south through west and returning froin south to north by east. Let GC be the direction of a horizontal nccdlc, so that CGC is the local attraction. Let x GC = % CG C = 8 x GC =% N G C =ự ô =W-4. Let w be the azimuth of thc principal section of the ship reckoned from G N towards the west, XGN -W=% - 0 M. Poisson arrives at the following equation, Conn. de Temps, 1841, p. 146, and Mem. de l'Institut, tom. xvi. p. 529. [A' cos 6 cos (4-W) + B cos 6 sin (4-w) + Csin ] sin (1.) = [D cos cos (y -w) + E' cos 6 sin ( w) + sin 0] cos 3 A', B, C, D, E, F arc constants, if therefore B с D E' F d, =b, and А? A л cos ( - )sins + csin (4-8) sin % + a tan 0 cos , = d cos (% - %) cos % + b sin (3-0) cos 3 + e tan 0 cos %. in which equation a, b, c, d, e are constants, which must be determined from observations at some one place, and which continue invariable as long as the positid of the iron in the ship replains unaltered. In order to do this thc local at. tractions corresponding to every azimuth of the ship’s head must be obtained and may be inserted in a table, or at least they must be found in sufficicnt number to afford any others by interpolation. The manner of obtaining such data from obis } .

·5 scized of going through the regular series on ship-board with more than usual diligence and care, so as to establish, by actual experiment in the only unexceptionable manner, the nature and amount of the corrections due to the ship's action for that particular geo- 'graphical position, and by the assemblage of all such observations to afford data for concluding them in general. Fourthly. No change possible to be avoided should be made in the disposition of considerable masses of iron in ships during the whole voyage; but if such change be necessary, it should be noted. servation is described in the small code of instructions which accompanies Mr. Barlow's plate, and is also given in the Nautical Magazine. Direction of ship’s head. Local attraction. Obscrved. 6 4. x ( C'. -0. XGC. < GN. CGC. N.B. Thir Table to be filled up from observation. If % -- (), - sin 8, d cosd, tetan 0 if = 180° - sin d cos da te tan ) if % = 90° - sin ez = c cos iz t atan d. if % · 270° sin 84 c cos 4 t a tano from these four equations, if d = mb, e = 9 ub, a, c, m and may casily be found, and b may be then obtained from any other known deviation. By solving cquation (1.) with respect to tan 5, the angle § 11y be found corresponding to the angle 5-8, d cos (% -8) + b sin (4-8) +etan tan cos (3-0) + csin ( 50)-ph a tau 0 Afterwards, by interpolation or reversion, the angle 5 - û may be obtained corresponding to the observed angle X, and a table of double entry formed, giving the local attraction for every value of 3, the arguments of the tablc being the obseropil angle ; and the dip. This table, according to the theory of M. Poi ought to continue available in all quarters of the globe so long as the disposition of the masses of iron in the ship remains unaltered. If when the ship's head is on thic magnetic north and south, no effects arise from ·6 1 Fifthly. When a traveller crosses the magnetic line of no dip it would be desirable to go through the observation for the dip with the instrument successively placed in a series of different magnetic azimuths, by which the influence of the ship’s magnetism in a vertical direction will be placed in evidence. On land, or on ice. If the completeness and excellence of the instruments with which the traveller is furnished authorize confidence if the results obtained by their use, the redetermination of the magnetic elements at points where they are already considered as ascertaincd, will be scarcely less desirablc than their original determination at stations where they have never before been observed. This is the more to be insisted on, as lapse of time changes these elements in some cases with considerable rapidity; and it is therefore of great consequence that observations to be compared should be as nearly cotemporary as possible, and that data should be obtained for eliminating the cffects of secular variations during short intervals of time, so as to enable us to reduce the observations of a series to a common epoch. On the other hand it cannot be too strongly recommended to the scientific navigator, studiously to seek every opportunity of landing . local attraction, as was the case in the experiments of Captaiu Flinders, y = 0, 2 = 1), and hence d= 0, e 0. If thic iron is symmetrically situate about the axis of the ship, then according to M. Poisson c = 0,2 = 0,6 = 0. cos (5-8) sin % + a tan 6 sin3 = b sin(5-8) cos . . (2.) Conn. de Temps, p.150. b sin ( 5-0) tans cos (2-1) + a tan * sin 3 {(1 - 1) cos a tan 0} sin ncarly. 1+1) vos* If 8' is the local attraction, the magnetic bearing of any conipass in any other distant part of the ship, or of the same compass after the disposition of the iron has been changed, cos (S' - %) sin % + atan 9 sin % = b'sin (3'-d') cos but % 8 hence cos (3-0) sins + a tan 6 sin %! b' sin (-) cos % eliminating tan (a' - a) sin X'sins tan ( a' sin % cos ab'sin costi From this equation it may be possible to compute a table of double entry, giving the local attraction without kuowing the dip, the arguments of the table being the obserred angles % of two compasses situate in different parts of the vessel. This table ought to continuc available so long as the disposition of the masses of iron remains unaltered. The two compasses must of course be so (listant as to have no sensible effect upon each other. The theory of Mr. Barlow's plate, according to M. Poisson, depends upon the practicability of so disposing the iron in the vessel as to give to the constants a and 6 in equation (2.) the particular values a = - 0,6 1.

·7 un points (magnetically speaking) unknown, and determining the elements of those points with all possible precision. Nor should it be neglected, whenever the slightest room for doubt subsists, to determine at the same time the geographical position of the stations *of observation in latitude and longitude. When the observations are made on ice, it is needless to remark that this will be universally necessary With this general recommendation, it will be unnecessary to enumcrate particular localities. In fact, it is impossible to accumulate too many. Nor can it be doubted that in the course of antarctic exploration, many hitherto undiscovered points of land will be encountered, cach of which will, of course, become available as a magnetic station, according to its accessibility and convenience. There are certain points in the southern regions which offcr great and especial interest in a magnetic point of view. These arc, first, the south magnetic pole (or poles), mcaning thereby the point or points in which the horizontal intensity vanishes and the needle tends vertically downwards; and secondly, the points of maximum intensity, which, to prevent the confusion arising from a double usc of the word poles, wc may provisionally term magnetic foci. It may herc bc mentioned, that no scientific datum of this description, nor any attempt to attain very high southern latitudes, can be deemed important cnough to be made a ground for exposing to extraordinary risk the lives of brave and valuable men. The magnetic polc, though not attained, may yet be pointed to by distinct and unequivocal indications; viz. by the approximation of the dip to 90°; and by thc convergence of the magnetic meridians on all sides towards it. If such convergence be observed over any considerable region, the place of the pole may thence be deduced, though its locality may be inaccessible. M. Gauss, from theoretical considerations, has recently assigned a probable position in lon. 146° E., lat. 66° S., to the southern magnetic pole, denying the existence of two poles of the same name, in either hemisphere, which, as he justly remarks, would entail the necessity of admitting also a third point, having some of the chief characters of such a pole intermediate between them. This may be made obvious without following out his somewhat intricate demonstration, by simply considering, that if a needle be transported from one such pole to another of the same name, it will begin to deviate from perpendicularity towards the pole it has quitted, and will end in attaining perpendicularity again, after pointing in the latter part of its progress obliquely towards the pole to which it is moving, a sequence of things impossible without an intermediate passage through the perpendicular direction. It is not improbable that the point indicated by M. Gauss will prove accessible ; at all events it cannot but be approachable sufficiently near to test by the convergence of meridians the truth of the indication; and as his theory gives within very moderate limits of error the true place of the northern pole, and otherwise represents ·8 the magnetic elements in every explored region with considerable approximation, it is but reasonable to hope that this point may be decided in Captain Ross's voyages. Should the decision be in the negative, i. e. should none of the indications characterizing the near vicinity of the magnetic pole occur in that region, it will be to be sought; and a knowledge of its real locality will be one of the distinct scientific results which may be confidently hoped from that Expedition, and which can only be attained by circumnavigating the antarctic pole compass in hand. The actual attainment of a focus of maximum intensity is rendered difficult by the want of some distinct character by which it can be known, previous to trial, in which direction to proceed, when after increasing to a certain point the intensity begins again to diminish. The best rule to be given, would be (supposing circumstances would permit it) on perceiving the intensity to have become nearly stationary in its amount, to turn snort and pursue a course at right angles to that just before followed, when a change could not fail to occur, and indicate by its direction towards which side the focus in question were situated. Another, and as it would appear, a better mode of conducting such a research, would be, when in the presumed neighbourhood of a focus of maximum intensity, to run down two parallels of latitude or two arcs of meridians separated by an interval of moderate extent, observing all the way in each, by which observations, when comparcd, the concavities of the isodynamic lines would become apparent, and perpendiculars to the chords, intersecting in or near the foci, might be drawn. Two foci or points of maximum total intensity are indicated by the general coursc of the lines in Major Sabine's chart in the Southcrn Hemisphere, onc about long. 140° E., lat. 47° S., the other more obscurely in long. 235° E., lat. 60° S., or thereabouts. Both these points are certainly accessible; and should the course of the navigator lead not far from each of them, they might be visited with advantage by a course calculated to lead directly across the isodynamic ovals surrounding them. Pursuing the course of the isodynamic lines in the chart above mentioned, it appcars that one of the two points of minimum total intensity, which must exist, if that chart be correct, may be looked for nearly abput lat. 25° Ss long. 120 W., and that the intensity at that point is probably the Icast which occurs over the whole globe. Now this point does not lie much out of the direct course usually pursued by vessels going to the Cape. It would therefore appear desirable to pass dircctıy over it, were it only for the sake of determining by direct measure the least magnetic intensity at present existing on the earth, an element not unlikely to prove of importance in the further progress of theoretical investigation. Excellent opportunities may be afforded for the investigation of all these points, and for making out the true form of the isodynamic ovals of the South Atlantic, both in bcating up for St. Helena, and in passing ·9 from thence to the Cape; in the course of which, the point of least intensity will, almost of necessity, have to be crossed, or at least approached very near. Nor is the theoretical line indicated by M. Gauss as dividing the northern and southern regions, in which free 'magnetism may be regarded as superficially distributed, undeserving of attention. That line cuts the equator in 6º east longitude, being inclined thereto (supposing it a great circle) 15°, by which quantity it recedes from the equator northward in going towards the west of the point of intersection. Observations made at points lying in the course of this line may hereafter prove to possess a value not at present contemplated. As a theoretical datum, the horizontal intensity has bcen recommended by M. Gauss, in preference to the total, not only as being concluded from observations susceptiblc of great precision, but as affording immediate facilities for calculation. As it cannot now be long before the desideratum of a chart of the horizontal intensity is supplied, the maxima and minima of this clement may also deserve especial inquiry, and may be ascertained in the manner above pointed out. The maxima of horizontal intensity art at present undetermined by any direct obscrvation. Tlicy must of necessity, however, lie in lower magnetic latitudes than those of the total intensity, as its minima must in higlier; and from such imperfcct means as we liave of judging, the situations of the maxima may be stated as occurring in 20° N. 80° E. I. 7 N. 260 E. II. 130 E. 10 S. 180 E. IV. Obscrvations have been made of the horizontal intensity in the vicinities of II. and III., and are decidedly the highest which have been observed anywhere. In general, in the choice of stations for determining the values of the three magnetic elements, it should be borne in mind, that the value of each new station is directly proportional to its remotcness from those already known. Should any doubt arise, thercfore, as to tlic greater or less eligibility of particular points, a reference to the existing magnetic maps and charts, by showing where the known points of obscrvation are must sparingly distributed, will decide it. 3 S. III. For such magnetic determinations as thosd above contemplated, the instruments hitherto in ordinary use, with the addition of Mr. Fox's apparatus for the statical determination of the intensity, will suffice; the number of the sca observations compensating for their possible want of exactness. The determinations which belong to the second branch of our subject, -viz. those of the secular changes, of the diurnal and other periodical variations, and of the momentary fluctuations of the magnetic forces,-rcquire, in the present state of ·10 our knowledge, the use of those more refined instruments recently introduced, and to be presently described. The variations to which the earth's magnetic force is subject, at a given place, may be classed under three heads, namely, 1. the irregular variations, or those which apparently observe no law; 2. the periodical variations, whose amount is a function of the hour angle of the sun, or of his longitude; and 3. the secular variations, which are either slowly progressive, or else return to their former values in periods of very great and unknown magnitude. The recent discoveries connected with the irregular variations of the magnetic declination, have given to this class of changes a prominent interest. In the year 1818 M. Arago made, at the Observatory of Paris, a valuable and extensive series of observations on the declination changes; and M. Kupffer having about the same time undertaken a similar research at Kasan, a comparison of the results led to the discovery that the perturbations of the needle were synchronous at the two places, although these places differed from one another by more than forty-seven degrees of longitude. This seems to have been the first recognition of a phenomenon, which now, in the hands of Gauss and those who are labouring with him, appears likely to receive a full elucidation. To pursue this phenomenon successfully, and to promote in other directions the theory of terrestrial magnetism, it was necessary to cxtend and vary the stations of observation, and to adopt at all a common plan. Such a system of simultaneous obscrvations was organized by Von Humboldt in the year 1827. Magnetic stations were established at Berlin and Freyberg: and the Imperial Academy of «Russia entering with zeal into the project, the chain of stations was carried over the whole of that colossal empire. Magnetic houses were erected at Petersburg and at Kasan; and magnetic instruments were placed, and regular observations commenced, at Moscow, at Sitka, at Nicolajeff in the Crimea, at Barnaoul and Nertschinsk in Siberia, and evenat Pekin. The plan of observation was definitely organized in 1830; and simultaneous observations were made seven times in the year, at intervals of an hour for the space of forty-four hours. In 1834 M. Gauss turned his attention to the subject of terrestrial magnetism; and having contrived instruments capable of yielding results of an accuracy before unthought of in magnetic researches, he proceeded to inquire into the simultaneous movements of the horizontal needle at distant places. At the very outset of this inquiry he discovered the fact, that the synchronism of the perturbations was not confined (as had been hitherto imagined) to the larger and extraordinary changes ; but that even the minutest deviation at one place of observation had its counterpart at the other. Gauss was thus led to organize a plan of simultaneous observations, (not at intervals of an hour, as before, but) at the short intervals of five minutes. These were carried on through twenty-four hours six* * Recently reduced to four.

·11 times in the year; and magnetic stations taking part in the system were established at Altona, Augsburg, Berlin, Bonn, Brunswick, Breda, Breslau, Cassel, Copenhagen, Kracow, Dublin, Freyberg, Göttingen, Greenwich, Halle, Kasan, Leipsic, Marburg, Milan, Munich, Naples, St. Petersburg, and Upsala. Extensive as this plan appears, there is much yet remaining to be accomplished. The stations, numerous as they are, embrace but a small ſvortion of the earth's surface; and what is of yet more importance, none of them are situated in the neighbourhood of those singular points or curves on the earth's surface, where the magnitude of the changes may be expected to be excessive, and perhaps even their direction inverted. In short, a wider system of observation is required to determine whether the amount of the changes (which is found to be very different in different places) is dependent simply on the geographical or on the magnetic co-ordinates of the place; whether, in fact, the variation in that amount be due to the greater or less distance of a disturbing centre, or to the modifying effect of the mcan magnetic force of the place, or to both causes acting conjointly. In another respect also, the plan of the simultaneous obscrvations admits of a greater cxtension. Until lately the movements observed have been only those af the magnetic declination, although there can be no doubt that the inclination and the intensity are subject to similar perturbations. Recently, at many of the German stations, the horizontal component of the intensity has been observed, as well as the declination; but the determination of another element is yet required, before we are possessed of all the data necessary in this most interesting research. The Magnetic Observatories about to be established in the British Colonies, by the liberality of the Government and of thc East India Company, will (it is hoped) supply in a great measure these desiderata. The stations are widely scattered over the earth's surface, and are situated at points of prominent interest with regard to the Isodynamic and Isoclinal lines. The point of maximum intensity in the northern hemisphere is in Canada; the corresponding maximum in the southern hemisphere is near Van Diemcn's Land; St. Helena is close to the line of minimum intensity; and the Cape of Good Hope is of importance on account of its southern latitude. Again, in India, Madras and Singaporc arc in the neighbourhood of the two lines of minimum intensity and of no dip, which in this region of the globe approach one another; and Simla, in the Himalaya mountains, is a station of interest and importance on account of its great elevation. At each observatory the changes of the vertical component of the magnctic force will be observed, as well as those of the horizontal component and declination; and the variations of thc two components of the force being known, those of the inclination and of the force itself are readily deduced. The simultaneous observations of these three elements will be made at numerous and stated periods, and there is every reason to hope that the directors of many of the European Observatories will take part in the combined system.

·12 But interesting as these phenomena are, they form but a small part of the proper business of an observatory. The regular changes (both periodic and secular) are no less important than the irreguli; and they are certainly those by which a patient inductive inquirer would seek to ascend to gencral laws. Even the empirical expro. - sion of the laws of these changes cannot fail to be of the utmost value, as furnishing a correction to the absolute values of the mag- *netic elements, and thereby reducing them to their mean amount. The hourly changes of the declination have been frequently and attentively observed; but with respect to the periodical variations of the other two elements, our information is as yet very scunty. The determination of these variations will form an important part of the duty of Magnetic Observatories; and from the accuracy of which the observations are susceptible, and the extent which it is proposed to give them, there can be no doubt that a very exact knowledge of the empirical laws will be the result. With respect to the secular variations, it might perhaps be doubted whether the limited time during which the observatories will be in operation is adequate to their detou nination. But it should be kept in mind that the monthly mean curresponding to each hour of observation will furnish a reparate result; and that the number and accuracy of the results thus obtained may be such, as fully to compensate for the shortness of the interval through which they are ful. lowed. A beautiful example of such a result, deduced from three years' observation of the declinati is to be found in the first ro). lume of Gauss's magnetical work, of which a translation is published in the fifth rimber of Taylor's Scientific Memoirs. Allusion has been made above to a different system of magnetic elements from that usually chosen. Before proceeding further, therefore, it is necessary to state inore fully what those elements are which have been taken as the immcdiate objects of research; and to describe the instruments which have been adopted for the purpose. The elements on which the determination of the Earth's Magnetic Force is usually based are, the declination, the inclination, and the intensity. If a vertical plane be conceived to pass through the di. rection of the force, that direction will be determined when its inclination to the horizon" is given, as well as the angle which thic plane itself forms with the meridian; and if, in addition to these quantities, we likewise know the number which expresses the rativ of the intensity of the force to some established unit, it is manifest that the force is completely determined. For many purposes, however, and especially in the delicate rescarches connected with the variations of the magnetic force, a different system of elements is preferable. The intensity being resolved into two portions in the plane of the magnetic meridian, one of them horizontal and the other vertical, it is manifest that these two components may be substituted for the total intensity and the inclina- 1 ·13 tion; while, at the same time, their changes may be determined with far greater precision. The former variables are connected with the latter by the relations X Rcos , Y = R sin ; in which R denotes the intensity, X and Y its horizontal and vertical components, and 8 the inclination; and the variations of 0 and R are expressed in terms of tlic variations of X and Y by the for mula: AY AX An= sin 201 Y X AR AX ΔΥ COSU + sinº R х Y As the instruments required for the observation of these clements are, for the most part, noveliu form, it will be useful to give a soniewhat detailed account of their construction and warious adjustments, before enteriug on the plan of observation to be pursued. DOKCLINITIO MAGNETOMETER. Construction.--. The essential part of the declination magnetometer is il magnet bar, suspended by fibres of untwisted silk, and inclosed in it box, tv protret it from thic agitation of the air. The bar is a rectangular parallelopedi, 15 inches in length, ths of an inch in breadth, and th of an in h 19 tluckness. In addition to the stirrup by which the bar is suspwed, it is furnished with two sliding pieces, one near each end. One of fresc pieces contains an achromatic lens, and the other a finely-divided scale of glass; tne scale being adjusted to the focus of the lus, it is mar -st na the apparatus forms a moving collimator, and that it: 10 lute position at any instant, as well as its changes of position from one instant to another, may be road off by a telescope at i distance. The aperture of the lens of this collimator is 1 inch, and its focal length about 12 inches. Each Hivision of the scale is şikth part of an inch, and the visual angle under which it is seen in the telescope is so considerable, that it may be readily subdivided into tenths by estimation. The correspouding angular quantity is about 43 seconds, and the readings may therefore be made to four seconds nearly. To the suspension tlıread is attuched a small cylindrical bar, the ends of which are of smuller diaineter, and support the stirrup which carries the magnct. The apertures in the stirrur by which it hangs un the cylinder, are of the form of inverted Yøs, so that the bearing points are invariable. A second pair of apertures at the other side of the magnet, serves for the purpose of inversal; and care has been taken to render the lines connecting the bearing points of each pair of Y's parallel, so that there inay be no difference in the amount of torsion of the thread in the two positions of the stirrup. The two pairs of apertures are at differeut distances from the magnet, in order that the line of collimation may remain nearly at the same ·14 height on inversal, and thus it may not be necessary to alter the length of the suspension thread. The stirrup, and the other sliding Pieces, are forined of gun metal. For the purpose of taking out the torsion of the suspension thread, the apparatus is furnished with a detorsion bar, which (with its aps pendages) is of the same weight as the magnet. It is a rectangular bar of gun-metal, furnished with a stirrup and collimator similar to othose of the magnet. A rectangular aperture in the middle receives a small magnet, the use of which is to impart a slight directive force to the suspended bar, without which the final adjustment of detorsion would be tedious and difficult. The frame-work of the instrument consists of two pillars of copper, 35 inches in height, firmly screwed to a massive marble base. These pillars are connected by two cross pieces of wood, one at the top, and the other 7 inches from the bottom. In the centre of the top piccc is the suspension apparatus, and a divided circle used in determining the amount of torsion of the thread. A glass tube (between this and the middle of the lower cross piece) incloses the suspension thread; and a glass cap at top covers the suspension apparatus, and complctes the inclosure of the instrument. The box is cylindrical, its dimensions being 20 inches in diameter by 7 inches in depth. It rests upon the marble slah, and encompasses the pillars; and it is so contrived as to be raised, when necessary, for the purpose of manipulation. There are two apertures in the box, opposite to each other. The aperture in front, used for reading, is covered with a circular piece of parallel glass, attached to a rectangular frame of wood which moves in dovetails; the pris- * matic error of the glass (if any) is corrected by simply reversing the -slider in the dovetails. The opposite aperture is for the illumination of the scale, and is furnished with a sliding piece of wood carrying a moveable mirror. In addition to the parts abovementioned, the instrument is provided with a second magnet, of the same dimensions as the first, to be used in measurements of absolute intensity; and a copper ring, for the purpose of checking the vibration *. * The declination magnetometers which were provided for the Antarctic expedition differ from that above described in one or two particulars, which liave been necessarily varied in order to enable the observer to read off very great deviations. The nagn has a second stirrup, which is furnished with a mirror, for the purpose of reading the declination chauges by reflexiou after the method of Gaussy The mirror is attached to a cross, which is fixed to the stirrup; it is kept in its place by three small screws, the heads of which project and hold it; the plane of the mirror may be adjusted by the movement of these scrcws. The deviation of the line of collimation (which, in this case, is the normal to the surface of the mirror) from the magnctic axis of the bar, is ascertained hy inversal, tlie stirrup being provided with a second pair of apertures, in the form of inverted Y's, for the purpose. There are two graduated scales to be used with the inirror. The first of these is a straight scalc, similar to that of Gauss, to be used for moderately large variations. This scale is 4 feet long, and the magnitude of cach division is .04082 of an inch. Every tenth division is numbered 1, 2, 3, &c. up to 100. The divisions are engraved and printed on paper which is attached to a slip of wood by sealing.

·15 Adjustment.-The instrument having been placed on its support, the base is to be levelled, and the whole then fixed in its place. The levelling of the base may conveniently be performed by the aid of plumb-line hanging in the place of the suspension thread; but no great precision is required in this operation, the chief object of which is that the suspension thread may occupy the middle of the tube, and that the magnet may be central with regard to its support. The suspension thread is then to be formed, and attached at one extremity to the roller of the suspension apparatus, and at the other to the small cylinder which is to bear the stirrup and magnet. Sixteen fibrcs* of untwisted silk are sufficient to bear double the load without breaking, and will be found to form in other respects a convenient suspension. These preparations being made, the adjustments are the following: 1. The sliders being placed on the magnet, the scalc is to be adjusted to the focus of the lens, and in such a manner that the centre of gravity of the sliders may be near thc middle of the bar. The adjustment to focus is readily made, by first adjusting the reading telescope accurately to focus by mcans of a star, and then using it to view the scale through the object lens of the collimator. The scale will be in the focus of the lens, when it is distinctly seen with the telescope. 2. The magnet is to be connected with the suspension thread by means of the stirrup, and to be moved in the stirrup until it assumes the horizontal position. This adjustment may be conveniently efſected by means of the image of the magnet, reflected from the surface of water or mercury, the object and its reflected image being parallel when the former is horizontal. The stirrup is bien fastened by its screws, and the magnet wound up to the desired height. As the thread stretches considerably at first, allowance should be made for this in the height. 3.† The magnet is then removed, and the unmagnetic bar (having its collimator similarly adjusted) is to be attached, without its sinali magnct, and allowed to swing for several hours. The bar having wax varnish. The second scale is graduated on a completc circle, 6 fect in diameter, and is to be used when the changes (regular or irregular) are excessive. Thc circle is of wood, and is in the shape of a flat ring, 9 inches in breadth, formed in segments so as to preserve its figure. At the exterior of the ring is a raised cdge (also built in segments), on the interior face of which the scale is attached. The scale itself is the same as the former; tlic nunbering of the divisions (every tenth) proceeding from 0 to 100, and then commencing anew. The aperture in the box is a rectangle, 6 inches long by 24 inches high, and is covered by a piece of wood sliding in dovetails, liaving a glass window of the usual size. The glass-cover will be used (with the collimator) when the variations are moderate; the uncovered aperture (with the mirror) when they are large, but not cxcessive ; and, when the variations are very great, the box niay be lifted altogether, so as to allow the entire range of azimuth. * Not the individual fibre of the silk-worin, but the compound fibre in the state in which it is prepared for spinning. + It is obvious that this step of the adjustment may precede the 1st and 2nd, where a saving of time is important.

·16 come to rest, or nearly so, its deviation from the magnetic meridian is to be estimated, and the moveable arm of the torsion circle turned through the same angle in an opposite direction. The plane of detorsion then coincides, approximately, with the magnetic meridian. 4. The magnet is then to be substituted for the unmagnetic bar, and the telescope being directed towards the collimator, the point of the scale coinciding with the vertical wire is to be noted when the magnet is in the direct and inverted positions. Half the sum of these readings is the point of the scale corresponding to the magnetic axis of the magnet bar; and half their difference (converted into angular measure) is the deviation of the line of collimation of the telescope from the magnetic meridian. The telescope should be moved through this angle in the opposite direction. 5. In order to take out the remaining torsion of the thread, the magnet is again to be removed, and the unmagnetic bar (with its small magnet attached) substituted. The deviation of this bar from the magnetic meridian should then be read off on its divided scale, and the moveable arm of the torsion circle turned through a given angle in the opposite direction. The deviation being again read, a simple proportion wik give the remaining angle of torsion; and the moveable arm being turned through this angle in the opposite direction, another observation will serve to verify the adjustment. The plane of detorsion then coincides with the magnetic meridian; and the magnet being replaced, the instrument is ready for use. Observations. The observations to be made with this instrument are, 1. of th“ absolute declination ; 2. of the variations of the declination; and 3. of the absolute intensily. For measurements of the absolute declination cach observatory is furnished with an azimuth instrument. This instrument being placed in the magnetic meridian of the declination instrument, the point of the scale coinciding with the wire of the telescope is to be observed ; the interval betweep this point and the point corresponding to the magnetic axis of the bar, converted into angular measure, is the deviation, D, of the line of collimation of the telescope from the magnetic meridian. The verniers of the horizontal circle being then read, the telescope is turned, and its wire made to bisect a distant mark, whose azimuth, Z, is known. If A denote the angle read off on the horizontal circle, it is mauifest that the angle between the magnctic and the astronomical meridians is A + 2 + D), Z and D being affected with their proper sigus. Some of the observatories are furnished with a transit instrument distinct from the theodolite; in others the transit instrument and theodolite are combined in one. In the former case the mark above referred to may be the central wire of the transit telescope, used as a collimator; so that for this case Z=0. In the latter case, the angle Z is to be previously determined by the belp of the transit instrument.

·197 0 The division of the scale corresponding to the magnetic axis of the bar is to be considered as the zero point, and must be determined with great exactness. It is obvious that this point will be given by the mean of two readings of the scale, with the magnet in the erect amil inverted positions, provided that care has been taken to eliminate the declination changes which may occur in the interval of the two parts of the observation. The obvious method of effecting this elimination is to determine the amount of the declination change, by means of an auxiliary apparatus, and to apply it as a correction to the second result. The same thing may, however, be effected by taking a series of readings of the declination instrument alone, with the bar alternately erect and inverted ; the time chosen for observation being one in which the declination changes are small and regular, and the readings being made in as rapid succession as possible. By comparing each result with the mean of the preceding and subseguent, and then taking the mean of all these partial means, a very accurate determination may be made. As an example of such a deterndination, we may take the following observation, made at the Dublin Magnetical Observatory, August 10, 1839. The first column of numbers contains the actual readings of the scale; the second the means of the preceding and subsequent readings; and the third the means derived from the combination of these last with the intermediate reading. I. Reversed 157.0 II. Direct .. 1932 Observations interrupted. III. Reversed 169.6 IV. Direct .. 182.0 168.8 175.4 V. Reversed 168.0 182.3 175.2 VI. Direct .. 1826 166:1 174.4 VII. Reversed 164.2 Combining the three means in the last column with the number 175.1 deduced from I. and II., the final mesin is 175.0. The following observation, made at the Dublin Observatory, on the same day, will illustrate the preceding method of determining the absolute declination. It must be premised that in this Observatory the transit telescope and the theodolite are distinct instruments. The centre of the theo. dolite is placed in the meridian line of the transit, and at the point where this line is cut by the magnetic meridian of the declination instrument; and the absolute declination is obtained by referring the telescope of the theodolite first to the line of collimation of the magnet, and then to that of the transit telescope. It will be seen from the preceding table that the mean of the read. с ·18 D ings III. V. VII., in the inverted position of the bar, which position in this instrument was the ordinary one, is 167.3; the mean time of observation being 1b 40m P. M., which corresponds nearly to the maximum declination of the day. But it has been already shown that the zero point of the scale is 1750; and the difference, 7.7, mui tiplied into 43.22 (the angle corresponding to a single division) gives, for the deviation of the line of collimation of the telescope from the magnetic meridian, 5' 321.8. The reading off on the limb of the theodolite was 150° 30'. The telescope of the theodolitc was then directed to the objectglass of the transit instrument, and its line of collimation brought into the astronomical nieridian by making the image of the vertical wire coincide with that of the middle wire of the transit. The read. ing of the limb was then found to be 358° 13' 50", and subtracting 180° (the magnet being to the north, and the transit telescope to the south of the thcodolite,) the true north corresponded to the angle 178° 15' 50" on the limb. The difference between this and the angle 150° 30', corresponding to the magnetic north, is A = 27° 45' 50". Hence, as in this case Z=0, the alsolute declination is A+D=27° 40' 17'.2. If, instead of the actual declination at any moment, we desired the mean declination of the day or of the month, we should employ (instead of the actual realing of the scale, 177.3) the corresponding mean result. Vye should thus obtain new valile of ), differing from the former by the amount of the declination change. In the observatories about to be estublished by Her Majesty's Government and by the East India Coinpany, the declination changes will be regularly observed with a fixed telescope, attached to a stune pillar, or to a firm pedestal of wood resting on solid masonry unconnected with the floor. Fiere then, instcad of referring the telcscope of the theodolite directly to the magnetic meridian by means of the moving collimator, the same result will be obtained, and probably in a better manner, by referring it to the line of collimation of the fixed:telescope. For this purpose it is only necessary to employ this telescope as a collimator, the instrument being reversed in its Y supports, if necessary. This mode of observation has the advantage of connecting the absolute determination directly with the regular series of observations; and it is manifest that it is sufficient, without any other means, to determine whether any, and what.changes may have occurred in the position of the fixed telescope. 2. In observing the variations of the declination the fixed telescope is alone employed. The observation consists simply in noting the point of the scale coinciding with the vertical wire, at three successive limits of the arc ot vibration. The three readings being denoted ·19 by a, b, c, the mean point of the scale corresponding to the time of the middle observation is 1 (u + 2b + c). his mode of observation is sufficient where the observer is not limited to a precise moment of observation. Otherwise the more exact me. thod pointed out by Gauss is to be preferred *. The changes of position of the scale may be converted into angular measure, the angle corresponding to one division being known. In general, however, this reduction will only be required in the monthly mean results. Before the truc changes of the declination can be deduced from the observed readings, it is necessary to apply a correction depending upon the force of torsion of the suspension thread. For supposing that the plane of detorsion has been brought (by the adjustments ahove described) to coincide with the mean magnetic meridian, it is manifest that on orery deviation of the magnet from its mcan pusition, the torsion force will be brought into play; and as this force tends to bring back the magnet to the mean position, the apparent deviations must be less than the truc. The ratio of the torsion force to the magnetic directive force is experimentally determined by turning the inuveable arın of the torsion circle through any given large angle (for cxample 90°), and observing the corresponding angle tlırough which the magnet is deflected. Let u denote the latter angle, and a the former; then the ratio in question is H F 11 in which H is the coefficient of the torsion force, and F the moment arising from the action of the earth's inagnetic force upon the free magnetism of the bar, tlic direction of the action being supposed to be perpendicular to its magnetic axis. The ratio of the two forces being thus found, the truc declination changes are deduced from the pparent, by multiplying them by the coefficient H I to F 3. The experiments necessary to determine the absolute intensity of terrestrial magnetism are of two kinds,-_experiments of deflection and experiments of vibration. The former of these give the ratio of the horizontal component of the carth's magnetic force to the moment of free magnetism of the deflecting bar, the latter determine the product of the same quantities. Deflection. The deflecting bar is to be placed at right angles to the magnetic meridian, and in the line drawn perpendicularly through the centre of the declination bar. Its centre is to be placed at * See Taylor's Scientific Memoirs, vol. ü. part v. p. 44. et seg. C2 ·20 m two different distances on this line; and in each position the observer is to note the angle of deflection produced by its action upon the suspended magnet of the declination instrument, the north end of the deflecting bar being placed, successively, towards the east and towards the west. Half the difference of the readings with the norta end east and west, converted into angular measure, is the deflection sought. The experiments are then to be repeated on the other side of the suspended magnet, and at the same distances, so that each angle of deflection is the mean of four observed results. Let r and gol denote the two distances, u and u' the corresponding angles of deflection. Also let X denote, as before, the horizontal part of the earth's magnetic force, and m the moment of free magnetism of the deflecting bar. Then the ratio of these quantities is given by the following formula : gali tanu got tan u X 2 (912 902) The quantity so obtained is to be corrected for the torsion force of the suspension thread of the declination magnetometer; this is H done by multiplying it by the number 1 + 7 In choosing the distances of the deflecting bar, we should take as the smaller that which produces a deflection nearly amounting to the entire limit of the scale, provided it be not less than four times the length of the,bar. The other should be greater than this in the ratio of 3 to 1, nearly. The distances themscles should be measured with great accuracy, and expressed in fuet und decimals of a foot. For the purpose of this measurement, each observatory is furnished with a standard scale and a beam coiupass. Vibration. The declination bar being now removed, the deflecting bar is to be placed in a temporary stirrup of silk or wire, and so attached to the cylinder of the suspension thread. Its time of vibration is now to be determined with accuracy, from at least 100 oscillations. This may be done by placing a fine mark (as a piece of fine silver wire) on the front end of the bar, and noting the time of passage across the fixed wire of the reading telescope. The telescope is for this purpose furnished with an additional object lens, which is to be placed over the other, so as to adapt the instrument to near distances. The time being observed, and the moment of inertia of the bar calculated, the product of the force of the earth and the moment of free magnetism of the bar is given by the formula 72 K T2 E mX in which in denotes the ratio of the circumference of a circle to its dia- ·21 17 meter, K the moment of inertia, and T the time of vibration. The moment of inertia, K, is given by the formula al + 62 K= M; 12 a and b denoting the length and breadth of the bar, and M its mass. In order to correct for the torsion force of the suspension thread, the time of vibration deduced from experiment must be multiplied by H H 1+ ; and therefore the square of the time by 1 + vi Again, inasmuch as these experiments necessarily occupy a considerable time, precision requires that we should apply a correction for the changes of the intensity which may occur in the course of the observation. In other words, we must reduce the observed value of m X (at the time of the experiment of vibration) to that which it had at the time of the experiment of deflection. This may at once be done by the aid of the horizontal force magnetometer, inasmuch as that instrument gives immediately the changes of m X. In this case, however,--that is, when there is a second magnet in the same room,—the deduced value of X will be (not the carth's magnetic force) but the resultant of that force and the force of the second magnet; and both experiment and calculation will be required to deduce the former. It is accordingly safer, perhaps, to remove the second magnet for the time to another apartment, and to determine the changes of the horizontal force by observing the time of vibration of this bar simultaneously with the two parts of the experiment above described. Thus, let t be the time of vibration during the experiment of deflection; t' the time during the experiment of vibration; then T" being the observed time, the corrected time of the deflecting bar is T=T'. m The values of X and of m X being known, that of X is at once m obtained by elimination. Let dent that X A, m X = B, then it is evi- X= EVE The number thus obtained for the force of the carth's magnetism expresses the ratio which that force bears to the unit of force,--the unit of force being that which, acting on the unit of mass, through the unit of time, generates in it the unit of velocity. These units are cntirely arbitrary; but for the sake of convenience in comparison, it is desirable that they should be the same in all the observations which shall be made according to this system. For the unit of mass, then, we may take a grain ; for the unit of time a second; and, if a foot be taken as the unit of space, the unit of velocity will be that of one foot per second.

·22 ... As the magnetic force operates effectively only on the free or uncombined elements of the magnetic fluid, we are to understand by the carth's magnetic force, its action on the clementary unit of free magnetism; and we must take for that unit the quantity of free mag. netism, which acting on another cqual quantity at the unit of distancc, exerts an effect equal to the unit of force already defined. The following example, taken from Gauss's memoir, Intensitas vis Terrestris ad mensuram absolutam revocata,' will serve to illustrate the preceding rules. The experiments were made in two apparatuses, denoted by the letters A and B, and with three bars, distinguished by the numbers 1, 2, 3. In the first place, the oscillations of bar 1 in the apparatus A, and of bar 2 in the apparatus B, were simultaneously observed. The time of a single oscillation, reduced to infinitely small arcs, was found to be for bar 1 158.22450, for bar 2 17.29995; the former being deduced from 305 oscillations, the latter from 264. The times were observer with a chronometer, whose daily rate was 145.24. Bar 3 bcing then suspended in the apparatus A, bar 1 was placed in the right line perpendicular to the magnetic meridian, both on the east and west side of the suspended bar (and in both cases with the north end east and west successively), and the deficction of bar 3 was observed in each position of bar 1. These experiments were repeated for two different distances, r, and with the following values of the deflection, u: y = 1.2 metres, u = 3° 42' 19.1.4 ; ri=1:6 1 34 19.3. During these experiments the oscillations of bar 2 were observed in the apparatus B. The time of a single oscillation, deduced from 414 oscillations, and corresponding to the mean time of the experiments of deflection, was 178.29484. H 1 The values of for bars 1 and 3, were found to be and 597.4 1 rexpectively. The moment of inertia of bar 1 was 721.6 K=4328732400, taking a millimetye and a milligramme as the units of distance and mass. Now, to calculate the results of these experiments :--the inequality of the times of oscillation of bar 2 proves that a slight variation of the terrestrial-magnetic intensity had taken place in the interval of the two parts of the observation. Accordingly, we have to reduce the observed time of oscillation of bar 1, to that corresponding to the mean state of the magnetic intensity during the second part of the observation. This time requires also a second reduction for the > ·23 chronometer's rate, and a third for the torsion of the thread. The reduced time is thus 17.29484 86400 T= 15.22450 X 598.4 X X =15.23580. 17.29995 86385.76 597.4 72 K ..mX= =179770600 = B. T% From the observed deflections we obtain pls tan u 5 tan du = 56528100; X 2 (w - g2) the millimetre being taken as the unit of distance; and correcting this quantity for the torsion of the thread, by multiplying it by the H 722.6 coefficient 1+ F 721.6 m m 56606437 = A.. X Accordingly, X= WBV 179770600 56606437 : 1.782088. The following Table contains the arc values of one division of the scale in each instrument, expressed in decimals of minute. a TABLE I. Designation of Instrument. Observatory. Arc value of one division. • • I. II. III. IV. V. A. B. C. H. H.M.S. Erebus Van Diemen's Land Montreal.. Cape of Good Hope St. Helena ... Simla Madras Singapore H.M.S. Terror 0.7267 0.7085 0.7208 0.7525 0.7108 0.6760 0.6746 0.6786 0.6786 HORIZONTAL FORCE MAGNETOMETER. The instrument employed in determining the horizontal component of the earth's magnetic force is similar, in principle, to the "bifilar magnetometer" of M. Gauss. It is a magnet bar, suspended ·24 by two equidistant wires, or (more accurately) by two portions of the same wire, the distance of whose bearing points is the same above and below; by the rotation of the upper extremities of the wire round their middle point, the magnet is maintained in a position at right angles to the magnetic meridian. It is manifest from the nature of this suspension, that the weight of the suspended body will tend to bring it into the position in which the two portions of the wire are in the same plane throughout. The moment of the directive force is G sin v; -v denoting the angle formed by the lines joining the bearing points above and below, or the deviation from the plane of detorsion; and G being expressed by the formula a2 G=W warning in which W denotes the weight of the suspended body, a half the interval of the wires, and I their length. The earth's magnetic force, on the other hand, tends to bring the magnetic axis of the bar into the magnetic meridian, with the force F sin u; in which u is thc deviation of the magnetic axis from the meridian, and F is the product of the horizontal part of the earth's magnetic force into the moment of frec magnetism of the bar. Te magnet being thus acted on by two forces, will rest in the position in which their moments are equal. When the instrument is so adjusted that u = 90°, or the magnet at right angles to the magnetic meridian, F=G sin v; and the ratio of the forces is known, when we know the angle v. But as one of these forces is constant, and the other variable, it is evident that the place of the magnet will vary around its mean position, and that the variations of angle are connected with the variations of the force. This connexion is expressed by the formula AF cotan v Au; the angle A u being expressed in parts of radius. Construction. The magnet bar is of the same dimensions as that of the declination instrument. The collimator, by which its changes of position are observed, is artached to the stirrup, and has a motion in azimuth *. The suspending wire passes round a small grooved wheel, on the axis of which the stirrup rests by inverted Y's; and the instrument is furnished with a scries of such wheels, whose diameters incrcase in arithmetical progression, (the common difference being about alth of an inch,) for the purpose of varying the interval of the wires. The exact intervals, corresponding to each separate * The horizontal force magnetometers belonging to H. M. ships Erebus and Terror are provided with a mirror attached to the stirrup of the magnet, and two scales, similar to those of the declination magnetometer. The mirror has a movement in azimuth, in the same manner as the collimator which it replaces.

·25 wheel, have been determined by the artist by gccurate micrometrical measurements; they are given in Tables III. and IV. in p. 30. The same interval is altered, at the upper extremity, by means of two screws (one right-handed and the other left-handed) cut in the same Cylinder; the wires being lodged in the intervals of the threads, and their distance regulated by a micrometer head. The interval of the threads of this screw (which is precisely the same for all the instruments) is Myths, or .02597 of an inch. The micrometer head is divided into 100 parts; and, as one revolution of the head corresponds to two threads of the screw, a single division is equivalent to .0005194, or the courth of an inch nearly. The micrometer head has been carefully adjusted by the artist, so that the index is at zero, when the interval of the wires is exactly half an inch. The collimator, in this instrument, is inclosed in a light tube attached to the stirrup. The aperture of the lens is about #ths of an inch, and its focal length about 8 inches. The divisions of the scale are the same as in the collimator of the declination magneto. meter; the corresponding arc values have been ascertained for each instrument by accurate experiment, and are given in Table II. p. 30. The larger parts of this apparatus,--the box, the framework, and the support,--are precisely similar to those of the declination magnctometer. In addition to the parts already described, the instrument is furnished with a spare magnet; a brass weight, required in determining the plane of detorsion of the wires relatively to the magnetic meridian; a thermometer, the bulb of which is within the box, for the purpose of ascertaining the interior temperature; and a copper ring used in checking the vibrations. Adjustments. The instrument being placed on its support, the base is to be levclled, and the whole apparatus fixed. Having then selected one of the small grooved wheels, and fixed it, temporarily, with its axis horizontal, the wire is to be passed round it; and the free extremities of the wire being passed through the corresponding holes in the suspension roller, placed beneath, weights are to be attached, and the two portions of the wire allowed to assume their natural position; the extremities may then be fastened to the roller, by introducing small wooden plugs in the holes. The parts are then to be inverted, and put in their proper places; the suspension apparatus resting on the divided circle, and the wire hanging down the tube. The collimator (its scale having been previously adjusted to focus*) is to be screwed on to the stirrup, and the latter attached to the axis of the grooved wheel by means of its Y's. The magnet is then introduced into the stirrup, and levelled ; and the wires wound upon the roller, until the collimator is at the desired height. These preparations being made, the adjustments are the following: 1. Determine experimentally the angle through which it is necessary to turn the moveable arm of the torsion circle, in order to deflect the magnet from the magnetic mcridian to a position at right angles * This adjustment has been already made by the artist.

·26 to it, the two positions being merely estimated. The cosine of this angle is, approximately, the ratio of the magnetic force to the tor- F sion forcc, or the value of the fraction The nearer this ratio is G to unity, the more delicate will be the instrument; practically, to will be found a convenient value. If, on making the foregoing experiment, the ratio should be found to fall below, or to exceed the proper limits, the torsion force must be altered by introducing a different wheel, and making the corresponding alteration in the interval of the upper extremities of the wires. 2. The magnetic axis being brought, approximately, into the magnetic meridian, by turning the moveable arm of the torsion circle, the collimator is to be turned, by its independent motion, until some point about the middle of the scale coincides with the vertical wire of the fixed telescope. This point of the scale is to be noted in the usual manner. 3. The magnet is then to be removed, and the brass weight attached. Note the new point of the scale which coincides with the wire of the telescope. Then, if the magnet had been placed in the previous experiment) in its direct position (i. e. north to north) the error of the plane of detorsion is (6 +1) v. ul being the difference of the two readings, converted into angular measure. If, on the other hand, the magnet had been reversed (i. e. north end tu south) the error is G -1 The moveable arm of the torsion circle is then to be turned through this angle, in the opposite direction; and the magnetic axis will be in the magnetic meridian. The difference of the two readings, corresponding to a given error, being much greater in the reversed than in the direct position of the magnet, it follows that the former affords a much more delicate method of making the desired adjustment. 4. The brass weight remaining attached, turn the moveable arm of the totsion circle through 90°. Then turn back the collimator, until some point about the middle of the scale coincides with the vertical wire of the fixed telescope ; and note the reading. 5. Now remove the brass weight, and replace the magnet. The magnetic force of the earth will bring it back towards the magnetic meridian, and the scale will be thrown out of the field of the telescope. Then turn the moveable arm of the torsion circle, until the point of the scale last noted is brought to coincide again with the wire of the telescope; the magnetic axis is then in the plane perpendicular to the magnctic meridian, and the adjustment is complete. The following adjustment, made at the Dublin Magnetical Observatory, will serve to exemplify the foregoing description.

·27 talo August 14.-Having attached the wire in the manner described, the wheel No.5, whose diameter=0.4607 of an inch, was introduced, and the stirrup and weight appended. The upper extremities of the wire having been adjusted to the interval of half an inch (the zero point) by means of the micrometer, the correction to be made in the upper interval was, in parts of an inch, 0.4607 - 0:5000 = -0.0393; and dividing this number by .0005194, the fraction of an inch corresponding to a single division of the micrometer head, the resulting number, 76, expresses the number of divisions through which the micrometer head was to be turned backward, so as to equalize the interval of the wires at top and bottom. The micrometer was accordingly turned back from 100 (or zero) to the division num. bered 24. 1. The magnet being now introduced into the stirrur, and the weight removed, it was found necessary to turn the moveable arm of the torsion circle through 151° 30', in order to deflect the magnet from the magnetic meridian to a position at right angles to it, the two positions being known only approximately. 'Hence the approximate value of the ratio of the forces was F = cos (151° 30') = .88. G 2. The moveable arm of the torsion circle was then turned, until the magnetic axis was, approximately, in the magnetic meridian : and the reading of the torsion circle was found to bo 260° 10'. The collimator was then turned, by its independent motion, until the division coinciding with the vertical wire of the fixed telescope was about the middle of the scale. The coinciding division was 140. 3. The magnet was now removed, and the brass weight attached ; and the new coinciding division was observed to be 118. The difference of the two readings is 22, and multiplying 1'.0776 (the arc value of a single division) by this difference, the angle v' = 241. Consequently (the magnet having been placed in the direct position) the error of the plane of detorsion was G 1.88 + V = 24' x 51'. F 0.88 Subsequent trials, however, showed that the error here deduced was greater than the truth; and that the two readings of the scale agreed, when the reading of the torsion circle was 260° 32'. 4. The brass weight being attached, the moveable arm of the torsion circle was turned through 90 degrees, so that its reading was 350° 32'. The collimator was then turned back, by its independent motion, until the coinciding division was about the middle of the scale. The reading of the scale was 134. 5. The brass weight was then removed, and the magnet replaced ; the scale was consequently thrown out of the field of the telescope. The moveable arm of the torsion circle was turned further, in the same direction as before, until the division of the scale last noted, ·28 134, again coincided with the wire of the telescope. The magnetic axis was then perpendicular to the magnetic meridian, and the ad. justment was complete. The reading of the torsion circle, after this step of the adjustment, was found to be 51° 2'; and consequently F v= 60° 30', and =sin v=8704. G Observations. The observations to be made with this instrument are those of the absolute value of the horizontal intensity, and its changes. From the explanation of the principle of the instrument, given above, it is manifest that it will serve to determine the moment of the force exerted by the earth upon the free magnetism of the suspended bar. Let X denote (as before) the horizontal part of the earth's magnetic force; m the moment of free magnetism of the bar; then m X=F, F having the same meaning as before (page 24). Hence, substituting the values of F and 6, we have a ·m X=W o quasi sin v; in which equation all the quantities of the second member may be obtained by direct measurement. The chief difficulty in this method consists in the determination of the quantity a, which should be known to a very small fractional part of its actual value. This difficulty has been overcome by the measuring apparatus connected with the suspension, which (as has been already stated) serves to determine the interval of the wires, at their upper extremity, to the quboth of an inch. The numbers given in Tables III. and IV. (p.30) for the lower interval, may be relied on to the same degree of accuracy. It is scarcely necessary to mention that the length of the wires, l, is to be mcasured between the points of contact above and below. The product of the earth's magnetic force into the magnetic moment of the bar being thus known, the ratio of the same quantities is to be determined by removing the bar from its stirrup, and using it to deflect the suspended bar of the declination instrument, according to the known method devised by Gauss. The experiments of deflection may, however, be performed without the aid of the second magnctometer, by operating upon another bar placed in the reverse position. This method has even the advantage in point of delicacy; but it labours under the disadvantage of requiring that the value of F should be determined for the second bar. The chief use of this apparatus is in observing the variations of the intensity. In these observations it is only necessary to note, at any moment, the point of the scale coinciding with the vertical wire of the fixed telescope, the mode of observing being precisely the ·29 same as in the other instrument. Let n be the number of divisions, and parts of a division, by which the reading at any moment differs from its mean value; then the corresponding variation of the angle (in parts of radius) is Au=na; a denoting the arc value in parts of radius) corresponding to a single division. Substituting this in the formula of page 24, and making k = a cotan v, we have AF AP= k n. The values of a have been determined for each of the instruments and are given in Table II. p. 30. Thus, in the instrument in use in the Dublin Observatory, the angle corresponding to a single division of the scale = 11:0776, so that a = 0003135. And since, in the adjustment of this instrument it was found that v = 60° 30', we have k .0003135 x cotan (60° 30') = '0001774. The quantity F, in the preceding formula, is the product of the earth's magnetic force into the moment offree magnetism of the bar; and, as the latter quantity varies with the temperature, it is necessary to apply a correction, before we can infer the true changes of the earth's force. This correction is easily deduced. Since F there is AF AX Am + F X Am so that the correction required is Let t denote the tempe- Xm, m m. m rature, in degrees of Fahrenheit; 9 the relative change of the magnetic moment corresponding to one degree; then Am = 9(-32). Accordingly, the changes of the earth's force will be expressed by the formula AX X kn + 9(t - 32). It is not necessary that these reductions should be applied to the individual results, except in cases of marked change, where it is desired to trace the progress of the actual phenomena. The results should be recorded as they are observed, in parts of the scale; and the reductions made in the monthly, or other mean values. The following Table contains the arc values of che division of the scale, in each instrument, expressed in decimals of a minute ; as also ·30 the same quantities reduced to radius, as the unit, by multiplying by the number .0002909. TABLE II. Arc values of one division. Designation of Instrument. Observatory. In Minutes. In parts of Radius. I. II. III. IV. V. A. B. C. H. H.M.S. Ercbus Van Diemen's Land Montrcal... Cape of Good Hope St. Helena .. Simla Madras Singapore H.M.S Terror 1.075 1.080 1.074 1.084 1.080 1.086 1:079 1.078 1.076 0003127 0003142 0003124 0003153 0003142 0003159 0003139 0003136 0003131 The following Tables contain the intervals of the axcs of the wires corresponding to each wlicel, in decimals of an inch; the wire used being that designated in commerce as "silver, finc 6." TALLE III. No. of Wheel. 1. 11.M.S. Erebus. II. TIT. IV. V. Van Diemen's Montreal. Cape of Good St. Flelena. Land. 1Iope. present 2 3 4 5 6 7 8 .2536 •3032 3:29 •4058 •4562 •5055 •5555 .6071 •2549 3959 .3516 •4088 •4555 5071 •5604 • 6071 -2529 3055 32.. 4078 .4581 •5042 *5588 • 6071 *2542 -3055 3497 • 4052 .4555 •5055 5565 .6097 .2536 .3063 .3513 -4071 •45-15 •5058 5591 6081 TABLE IV. A. No. of Wheel. B. Madras. Simla. C. Singapore. H. H.M.S. Terror. 7 8 9 10 11 12 .3614 •4153 4620 •5182 •5663 .6153 3613 4159 4643 •5146 •5659 .6111 3598 *4124 4663 5182 •5666 •6120 3637 4146 4640 •5156 5672 .6143 ·31 יין VERTICAL FORCE MAGNETOMETER. The instrument used in determining the changes of the vertical component of the magnetic force is a magnetic needle resting on agate plancs, lsy knife-udges, and brought to the horizontal position by weights. l'rom the changes of position of such a needle, the changes of the vertical force may be inferred, when we know the meun inclination at the place of obscrvation, and the times of vibration of the necdlc in the vertical and in thc horizontal planes. For. it may be shown that AF cotan 0 ani where A y denotes the change of the angle in parts of radius, 0 the inclinatiuit, and T and T the times of vibrntion of the needle in the cerlicul and horizontal planos respectively. l'onstruction. The magnetic weedle is 12 inches in length. It has a cross of wires at each extremity, attached by means of a small ring of copper; the interval of the crosses being 13 inches. The axis of the ricodle is formed into a knife edge, the edge being adjusted to pass as nearly is possible through the centre of gravity of the unloaded instrunicnt. The Wiglits by which the other adjustments are effected arc small brass crews moving in fixed nuts, one on cach arm ; the axis of one of the scrows being parallel to the magnetic axis of the needle, iuud 11.1 of the other perpendicular 10 it. The gate planes upon which the needle rests are attached to a solid support of copper, which is firmly fixed io a massive marble basc. in this support there is a provision for raising Se needlc off the planes, the contrivance for cffecting this object being similar to that employed in the inclination instrument. The whole is covered with an oblong hox of mahogany, in one side of which are two small glazcil apertures, for the purpose of reading; the opposite side of the box is (nyired with plate glass. thermom.. ter, within the box, sluws the temperature of the interior ajr; and a spirit level, attached to the inarble base, serves to indicate any change of level which may occur in the instrument. The positious of the needle at any instant is observed by means of two micrometer microscopes, one opposite each end. These microscopes are supportel on short pillars pf copper, attached to the base of the instrument. They are so adjusted that one complete revolution of the micronicter screw corresponds to about 5 minutes The micronieter head is divided into 50 parts; and, consequeutly, the arc corresponding to a singlesdivision is about 0'1. In addition to these parts, the apparatus is provided with a brass bar of the same length as the magnet, (furnished, like it, with cross wires at the extremities, and knife-edge bearings,) for the purpose of determining tlic zero points of the micrometers; a brass scale, divided to 10', used in ascertaining the value of the micrometer divisions ; and a horizontal needle, to be employed in determining the azimuth of the vertical plane in which the needle moves. of are.

·32 Adjustments.-The following are the adjustments required in this instrument : 1. The instrument being placed on its support, in a suitable position with respect to the other two instruments, the azimuth of the plane in which the needle is to move may be adjusted in the follow. ing manner. The plane is made to coincide, in the first instance, with the magnetic meridian, by means of the horizontal needle which moves upon a pivot fixed to the top of the scale. A small theodolite (or other instrument for measuring horizontal angles) is then placed on the base; and its telescope brought to bear on a distant mark. The telescope should then be moved through a horizontal angle equal to the intended azimuth of the instrument, but in an opposite direction. The base of the instrument is next to be turned, without disturbing the theodolite, until the mark is again bisected by the wire of the telescope: it is then in the required azimuth. The base should then be levelled, and permanently fixed. 2. The fixed wires of the microscopes are then to be adjusted to the same horizontal line. This is effected by means of the brass needle. This needle being placed upon the agate planes, by its knife-edges, and allowed to come to rest, it is manifest that the line joining the cross wires will be horizontal, provided it be perpendicular to the line joining the centre of gravity and the axis. To effect this latter adjustment, the needle (a great part of whose weight is disposed below the knife-edge) is furnished also with a small moveable weight. The test of the adjustment is similar to that of the corresponding adjustment of the ordinary balance. The moveable wire of one of the microscopes being brought to biscct the cross, if the adjustment is complete, it will bisect the cross at the other extremity upon reversal; if not, the position of the needle will indicate in what manner the weight is to be moved. A horizontal line being thus obtaincd, the fixed wires of the microscopes are to be adjusted to it, by moving the capstan-headed screws with which they are connected. 3. The last adjustment is that of the magnetic needle to the horizontal position. To effect this adjustment the needle is furnished with two moving weights, one on each arm. These weights (it has been already stated) are screws moving in fixed nuts, one in a direction parallel to the magnetic axis of the needle, and the other in a direction at right angles to it. By the movement of the former the needle is brought to the horizontal position; and by that of the latter, the centre of gravity is made to approach the centre of motion, and the sensibility of the instrument thereby augmented. It is manifest that the line joining the two crosses at the end of the needle will, in general, deviate from the magnetic axis. The amount of this deviation may be ascertained (before the instrument is permanently fixed) by reversing the needle on its supports, the plane of the instrument being perpendicular to the magnetic meridian. Half the difference of the readings, in the two positions, will be the deviation sought. Observations. In observing the variations of the vertical force ·33 with this instrument, it is only necessary to bring the moveable wire of each micrometer to bisect the opposition cross of the needle The interval between the fixed and moveable wires, expressed in angular measure, is the deviation of the needle from the horizontal position; and the changes of the vertical force are thence obtained by multiplying by a constant coefficient. If n denote the number of divisions, of the micrometer head, corresponding to the interval of the wires, Δη Ξηα, a denoting the angle (in parts of radius) corresponding to a single division. Consequently, the changes of the force are expressed (as in the case of the other component) by the formula AF =kn; F in which the constant coefficient is T2 k = a cotan 0. T2 The quantity F in the preceding formula is the product of the vertical component of the earth's magnetic force multiplied by the moment of frec magnetism of the needle; or F = m Y. accordingly the results thus deduced require a correction for the effects of tenperature upon the quantity m. This correction is simila. to that applicd to the horizontal intensity, and the corrected expression of the changes of the vertical component is conscquently AY = kn + 9 t - 32); Y where t denotes the actunl temperature (in degrees of Fahrenheit) at the time of obscrvation, and q the relative change of the magnetic moment of the needle corresponding to one degree. As in the case of the other instruments, however, it is not in general necessary to apply those reductions to the individual results. AY TIMES OF OBSERVATION. The objects of inquiry in Terrestrial Magnetism may be naturally classed under two heads, according as they relate, 1. to the absolute values of the magnctic elements at a given epoch, or their mean values for a given period; or 2. to the variations which these elements undergo from one epoch to another. It will be convenient to consider separately the observations relating to these two branches of the subject. Absolute Determinations. By the method of observation which has been suggested for the absolute declination, every determination of the position of the declination Bar is rendered absolute. We have only to consider the varying angle between the magnetic axis of the bar and the line of collimation of the fixed telescope, as a correction to be applied to the D ·34 constant angle alrearly determined) between the latter line and the meridian. It is manifest that if the fivity of the line of collimation of the telescope could be depended on, a single determination of the latter angle would be sufficient. But this is not to be trusted for any considerable period; and it will be therefore necessary, from time to time, to refer the line of collimation of the telescope to the meridian, in the manner already explained. This observation may be repeated once a month, or more frequently if any change in the position of the telescope be suspected. In the case of the intensity, there is another source of error (beside that due to a change in the position of the instruments) which can only be guarded against by a repetition of absolute measurements. The magnetic moment of the magnet itself may alter; and the observations of intensity changes afford no means of separating this portion of the effect from that due to a change in the earth's magnetism. This separation can only be effected by means analogous to those employed in the determination of the absolute value of the horizontal intensity; and accordingly one or other (or both) of the methods proposed for this determination should be occasionally resorted to. It is desirable that this observation should be repeated once in every month; and more frequently, whenever the changes observed with the horizontal force magnetometer indicate, by their progressive character, a change in the magnetic moment of the suspended bar. It would be easy, in theory, to devise a method by which the vertical force magnetometer might be made to serve in determining the absolutc value of the vertical intensity. The means which at present offer themselves appear, however, to be surrounded with practical difficulties; and it seems safer to deduce this result indirectly. From the formulæ given in page 1.2, we have Y = X tau 0); so that if the inclination, 0, be known, and the horizontal intensity, X, determined in absolute measure, the vertical intensity, Y, is inferred. For the purpose of observing the element o, each observatory is furnished with an inclination instrument. The observation should be made in an open space, sufficiently remote from the magnets of the observatory, and from other disturbing influences; and a series of measures should be trken simultaneously with the two intensity magnetometers, for the purpose of eliminating the changes of the inclination which may occur in the course of the observation. As to the mode of observation, the best seems to be the usual one, the plane of the circle coinciding with the magnetic meridian; but for the purpose of testing the axles of the needles, and the limb of the instru. ment, it is desirable that a series of observations should be made in various azimuths,--for example, every 30° of the azimuth circle commencing with the magnetic meridian. The inclination is then inferred, from each pair of corresponding results, by the formula cotan? 0 cotanº, + cotan?'; y and n' being the observed angles of inclination in two planes at right ·35 angles to one another. Where the inclination is great (as in Canada), this method will serve to test only a limited portion of the circumference of the axle and limb. In this case the best course appears to be that pointed out by Major Sabine *, namely, to convert one of the needles, temporarily, into a needle on Mayer's principle, by loading it with sealing-wax; and to deduce the inclination, from the angles of position of the loaded needle, by the known for. mula of Mayer. The observations here suggested having been very carefully made, and the inclination changes eliminated in the manner above explained, the observed difference between the mean and the result obtained in the magnetic meridian, should be applied as a correction for the errors of axle and limb to all future observations made in the meridian. These observations should be made at the same periods as those of the absolute horizontal intensity. Variation of the Elements. . The variations of the magnetic elements are, 1. Those variations whose amount is a function of the hour angle of the sun, or of his longitude ; and which return to their original values at the same hour in successive days, or the same scapon in successive years. These, from their analogy to the corresponding planetary inequalities, may be denominated periodical. 2. The variations, which are either continually progressive, or else return to their former values in long and unknowe periods; these may in like manner be deno. minated secular. 3. The irregular variations, whose amount changes from one moment to another, and which observe (apparently).no law. The periodical variations (with the exception of those of the declination) have hitherto been littlc studied; and, even in the case of the single element just mentioned, the results have scarcely gone beyond a general indication of the hours of maxima and minima, and of the changes of their amount with the season. The subject is nevertheless of the highest importance in a theoretical point of view. The phenomena depend, it is manifest, on the action of solar heat, operating probably through the medium of thermo-electric currents induced on the earth's surface. Beyond this rude guess, however, nothing is as yet known of the physical cause. It is even still a matter of speculation whether the solar influence be a principal, or only a subordinate cause, in the phenomena of terrestrial magnetism. In the former case, the periodical changes are to be regarded as the effect only of the variations of that influence; in the latter, they must be considered as its entire result, the action in this case only serving to modify the phenomena due to some more potent cause. It may be fairly hoped that a diligent study of this class of phenomena will not only illustrate this and other doubtful points in the physical foundation of the science; but also, whenever that physical cause shall come to be fully known, and he made the basis of a mathematical theory, the results obtained will serve to give to the latter a numerical expression, and to test its truth. Even Reports of the British Association, vol. vii. p. 55. D 2 * ·36 the knowledge of the empirical laws of the hourly and monthly fluctuations must prove a considerable accession to science; and (as one of its more obvious applications) will enable the observer to reduce his results, as far as this class of changes is concerned, to their mean values. For the complete determination of the hourly and monthly changes of the magnetic elements, a persevering and laborious system of observation is requisite. The irregular changes are so frequent, and often so considerable, as (partially at least) to mask the regular; and the observations must be long continued at the same hours, before we can be assured that the irregularities do not sensibly affect the mean results. Again, in a theoretical point of view, the nocturnal branch of the curves by which the periodical changes are represented is quite as important as the diurnal; and it is manifest that nothing can be done towards its determination without the cooperation of a number of observers. At cach of the Observatories about to be founded by Her Majesty's Government and by the East India Company, there will be three assistant observers placed under the command of the Director ; and it is intended that the observations shall be taken every two hours throughout the twenty-four. In order that this series of observations, which is especially destined for the determination of the periodical changes, may at the same time cast some light upon the irregular movements, it is proposed that they shall be simultancous at all the observatories. The hours which have been agreed upon are the even hours (0, 2, 4, 6, &c.) Göttingen mean time. It is likewise intended that one observation of the twelve shall be a triple observation, the position of the magnets being noted according to an arrangement which will be hereafter explained. The time of this triple observation will be 2 P.M., Göttingen mean time. The barometer, and the wet and dry inermometers, will be registered at each of the twelve magnetic hours. No observation will be taken on Sunday. No distinct series of observations is required for the determination of the secular variations. In the case of the declination, the yearly change will be obtained by a comparison of the monthly mean results (for the same month and same hour) in successive years. The observations of two years only will thus furnish 144 separate results, from which both the periodical and the irregular changes are eliminated; so that great precision may be expected in the final result, notwithstanding the limited period of observation. The same mode of reduction will apply to the two components of the intensity, provided that no change shall have taken place in the magnetic moment of the bars employed. In the latter event, recourse must be had to the absolute determinations for a knowledge of the secular changes. The subject of the irregular movements has acquired a prominent, and almost absorbing interest, from the recent discoveries of Gauss. It has been ascertained that the resultant direction of the forces, by which the horizontal needle is actuated at a given place, is incessantly varying, the oscillations being sometimes small, sometimes very con- ·37 siderable ;-that similar fluctuations occur at the most distant parts of the earth's surface, at which corresponding Observations have been as yet made ;—and that the instant of their occurrence is the same everywhere. The intensity of the horizontal force has been found •subject to analogous perturbations. For the full elucidation of the laws of these interesting phenomena, it is of the first importance that the stations of observation should be separated as widely as possible over the earth's surface, and that their positions should be chosen near the points of maxima and minima of the magnetic elements. This has been in a great measure accomplished as regards the observatories about to be founded by Her Majesty's Government and by the East India Company. The stations are wide asunder in geographical position, and they arc in the neighbourhood of points of prominent interest in reference to the isodynamic and isoclinal lincs. The results of observation at these stations will soon testify whether the shocks to which the magnetic needle is subject, are of a local or of a universal character as regards the globe; and in either event we may expect that they will furnish information of great value (in reference to a physical cause) as to thc magnitude of thc phenomena in different places, and the elements on which it depends. In the observations destined to illustrate this class of phenomena, it is proposed to follow the plan laid down by Gauss, as nearly as the proposed extension of that plan will permit. The Göttingen terms * will be observed as agreed on by the German Magnetic Association; but in place of confining them to four in the year, as at present, they will be extended to twelve. The four terms at present agreed upon by the German Magnetic Association occur in the months of February, May, August, ando November, and commence on the Friday preceding the last Saturday of the month, at 10 P.M. Göttingen. mean time. It is proposed that the eight additional terms shall be held on the Wednesday • ncarest the 21st of the eight remaining months; the hour of commencement being the same as before. Thus, in the course of the next three years, there will be thirty-six terms of simultaneous obscrvation. The days of commencement of these terms are given in the subjoined table. The hour of commencement will be, in all cases, 10 P.M. Göttingen mean time. In these terms the declination magnetameter will be observed at intervals of five minutes, the corresponding moments of observation being the full minutes, 0, 5, 10, 15, &c. Göttingen mean time. Each of the force magnetometers will be obsexved at intervals of ten minutes ; the epochs of observation of the horizontal force magnetometer being 2m 30%, 12m 30%, 22m 30$, &c., and those of the vertical force magnetometer 7m 30%, 17 m 30%, 27m 30%, &c. Göttingen mean time. By this arrangement an observation will be taken every 2 minutes, and the observer will have sufficient time between the observations to transfer his attention from one instrument to the next without embarrassment or confusion. * For the details of the arrangements of the "term observations," see the translation of Gauss's memoir in Taylor's Scientific Memoirs, vol. ii. part v.

·38 Days of commencement of the terms of simultaneous observation, during the years 1840, 1841, and 1842. Month. 1840. 1841. 1842 June. January Wednesday 22 Wednesday 20 Wednesday 19 February Friday 28 Friday 26 Friday 25 March Wednesday 18 Wednesday 24 Wednesday 23 April Wednesday 22 Wednesday 21 Wednesday 20 May Friday 29 Friday 28 Friday 27 Wednesday 24 Wednesday 23 Wednesday 22 July .. Wednesday 22 Wednesday 21 Wednesday 20 August. Friday 28 Friday 27 Friday 26 September Wednesday 23 Wednesday 22 Wednesday 21 October Wednesday 21 Wednesday 20 Wednesday 19 November Friday 27 Friday 26 Friday 25 December.... Wednesday 23 Wednesday 22 Wednesday 21 In case of the occurrence of Auroras, the hourly should at once be exchanged for uninterrupted observation, should that not be actually in operation. The affections of the magnetometers during thunderstorms, if any, should be noticed, though it is at present believed that they have no influence. During an earthquake in Siberia in 1829, the direction of the horizontal needle, carefully watched by M. Erman, was uninfluenced; svould a similar opportunity occur, and circumstances permit, it should not be neglected. 2. FIGURE OF THE EARTH. If the traveller be provided with invariable pendulums, with all the necessary apparatus for determining the length of the seconds pendulum, it will be highly desirable to have this important observation made at several points, especially in high southern latitudes, and generally speaking at points as remote as possible from those at which it has already been determined. The selection of these must depend on local circumstances, as regards convenience for landing the instruments and executing the operations, as well as on the times of arrival at the several points. It would also be desirable, if a convenient opportunity occurs, to swing the pendulums on the top of some high mountain; in which case they should also be swung at the foot of the same mountain, in order to determine the difference produced by the elevation, or other effect of the high land. Another experiment which it would be desirable to make, is to swing the pendulum on a large field of fixed ice, as far from the land as possible; and likewise on the nearest shore to such position. In all these cases more than one pendulum should be used; and ·39 at least three knife edges should be employed, in order to guard against any unforeseen anomaly that may arise. It is scarcely necessary to state, that the direction of the line of motion, with respect to the magnetic meridian, should be noted at each station, 3. TIDES. Our knowledge of the tides is so incomplete that there are few coasts on which good tide observations will not be valuable, especially if they are made with a due apprehension of the purposes to which they should be directed. One of these purposes is to determinc the manner in which the tide travels from one part of the occan to another; and in order to answer this purpose the observer must ascertain the establishment, or time of high water with reference to the moon's transit. The interval at which the high water follows the transit of this luminary is not constant; it varies to the annount of about eighty minutes in the course of a semilunation, according to the law expressed below in the Table of the Semimenstrual inequality, p. 43; and hence the corrected establishment, or lunitidal interval, as also the vulgar establishment, or time of high water at new and full moon, may be determined approximately by observations made at any period of a lunation : but 11o deterinination can be very exact which does not include several days' observations. The traveller or navigator who would contribute towards our knowledge of the inotion of the tide ought to make tide observations at several places along the same line of coast, so as to ascertain the establishment at cach place, and thus to determine in what direction and at what rate the tide travels. This it is desirable to have done on all coasts. In islands it is especially desirable to have observations on opposite sides of the island, so as to ascertain what part of the coast is first reached by the tide and what last. Great variations will often be found to occur in the range of the tide (namely, the range from low water to high water) even in a line of coast of no great extent. In order that these variations may clearly come into view, it is proper to observe low waters as well as high waters. Another purpose for which tide observations are necessary is to determine the laws of the tides at a given place. The laws of the tides, as applicable to practical uses, are commonly assured to be nearly the same at all places. If this were the case, all that it would be necessary to know, in order to obtain the laws for any particular place, would be the establishment of the place, and the range (greatest, least, or inean) of the tide. In fact, however, it is found that the senimenstrual inequality of the times varies considerably from place to place, and that the semimenstrual inequality of the range is not proportional to the range itself. Hence it is desirable to make continued observations at several places, so as to compare their semimenstrual inequalities. It is probable alsс that other corrections (as those for lunar and solar parallax and declination) differ at different places in the same manner. Hence tide tables con- ·40 structed for one placeand applied to another place by merely altering the establishment, will often be found very inaccurate. There is one feature of the tides which, though nearly universal, has only been lately made a subject of observation, and which in many cases extremely modifies the phenomena. This is the diurnal' inequality, or difference of the two tides on the same day. This difference affects both the heights and the times; commonly the former most; and is often so large as greatly to disguise the phenomena of the tides. In extreme cases indeed (when its effect either on the time or on the height is very great) it produces the appearance of one tide only in twenty-four hours. At places where opportunity occurs it would be very desirable to make observations which may determine this inequality; for we are at present actually ignorant whether a considerable inequality of alternate tides be the rule or the exception. Observations continued for a fortnight at any place would ascertain the existence of this inequality, but it would be necessary to observe for a considerably longer period in order to determine its amount and law with exactness. It is to be recollected that the inequality disappears once every fortnight, and changes from me tide to another when the moon passes from south to north, or froin north to south declination. As to the mode of making observations; when the water is still, its height, and the time of its reaching the greatest height (or the least) may be observed by means of a pole carefully divided into feet and inches, and fixed upriglit in the water; or by any cquivalent contrivance. The graduations of the pole, if fixed, should be read from below upwards, and should commence or have their zero at a point so far below the level of low water as to admit of no pos- "sibility of the surface of the sea ever descending so far. By this means all the readings are positive, and ali doubts and consequent inevitable mistahes of the + and rcadings are obviated; the observed range of the tide is then in all cases the difference of the greatest and least semidiurnal readings--the mid-water point half their sum. The mean of all the mid-water points, in a long series of observations, gives the division of the scale corresponding to the mean level of the sea. In a short series the correction will be rcquired for this, which, in the case of a single day's observations, may be considerable. The heights registered however should be those rear off on the scale, and not those deduced therefrom by the subtraction of any assumed constant for the mean sea-level. A much more exact and preferable mode of observation, howcver, is to use an upright tube or chamber, in which is a float carrying an upright measuring rod carefully graduated into feet and inches from above downwards (the zero of the scale being above the top of the rod, so that it would require to be prolonged, if it were required to read off the zero). This rod should work through a loose collar near the point where the height is read off (which must be done by a fixed index), and friction rollers may be applied if the float have but little weight and buoyancy. The tube or chamber should not be less than three or four inches in diameter, smooth on ·41 the inside, and giving access to the water by small opening about 4 inch diameter ncar the bottom, but not so near as to allow the entry of sand, &c., and guarded from that of sea weed, or sea animals, by a grating. This contrivance will ncutralize the effect of 'the waves on the float. The time should be observed by a watch or clock properly regulated to mean solar time at the place. If it strike minutes, so much the better. Some observers have noted, as the time of high water, the time when the water begins to fall, which is manifestly, as a general rule, too late. Others when it ceases to rise, which taken as a rulc, would give a result too early, and both liable to serious mistakes in irrcgular tides. If the observer follow cither of these inaccurate methods, he should state which. The only exact method is for the observer to be on the watch at least a quarter of an hour before the expected time of high water, and to note the reading of the float at every entirc minute from the instant he arrives to that when he desists; and his obscrvations should be continued for not less than ten minutes after he begins to feel assured that the high water is past. If provided with a supply of engraved squares, he may then (at leisure or while the observations are proceeding) take the time in minutes for an abscissa and the readings in inches for ordinates, project the curve of high water,-round off its angles so as to destroy accidental jerks, (which will require a little practice and the exercise of a discreet judgement,) and both the time and height of high water will be at once indicated by the course of the curve with all the precision the observation is capable of. But should he not feel competent to this operation, (which however a few trials renders very easy,) he may select for one observed height when the water has been certainly rising, a corresponding equal height when it has been certainly falling, and take the mean between these for a time of high water. And the time being thus concluded for cach observed rising height, to which a corresponding equal sinking height has been noted, the mean of all the times of high water so deduced will be to be taken for the true time. In all cases the original notes of the minutely observations should be preserved. In must be recollected, however, that observations made in this manner must not be directly compared with those in which the surface ccasing to rise or beginning to fall is observed; allowance must be made for the time during which the surface is apparently stationary. The height of the zero point of the readings, above some fixed and probably permanent mark or linc, (which should be described accurately and circumstantially so that it may be easily recovered) should be carefully ascertained. When the graduations are on a fixed pole, it is necessary for this purpose to level from the mark pitched upon, to any one (stated) of the lines of graduation. Thus, if the graduation of twenty feet is found to be eight feet above the fiduciac point chosen, the zero of the scale, were it prolonged so far downwards, would be twelve feet below it. When a float is used, it ·42 is the difference of level between the index by which the reading off iş made and the permanent mark, which must be ascertained. Each obscrvation should carry the observer's name or initials ; and the time in which the observations are made, whether inean or apparent solar time, as also the clock's error, and how obtained, should be carefully stated. Mean time as above recommended is now universally prevalent. Low water must be observed in the same manner (mututis mutandis) as high. In cases where, from the slope of the shore and other causes, low and high waters cannot be observed at the same station, a levelling operation will be necessary to bring to exact coincidence the zeros of the two graduated scales which must then be necessarily employed. Mr. Palmer has described in the Philosophical Transaétions a self-registering nachine, which is intended to give the time of high water. Such a machine is in operation in Her Majesty's dock-yard at Sheerness, and is riescribed in the Nautical Magazine for October 1832. The principle consists in a style or pencil which is moved horizontally by the tide along the summit of a cylinder which is turned round slowly and uniformly; the pencil describes a curve upon paper wound round the cylinder, which curve indicates the fluctuation of the water. The motion of the tide being originally vertical is changed by a very simple mechanical contrivance. Another contrivance destined for the same purpose has been invented by Captain Lloyd. The following is the form of registering tide observations employed under the direction of the Admiralty, in which it will be observed are noted, 1. The time and height of high water; 2. The tinie and height of low water; 3. The direction and force of the wind. Register of the Tides at Portsmouth Dock-yard. High Water. Low Water. 1834. Wind. Range of Tide. January. Time. Hicight. Time. IIeight. Direction. Force. N.N.W. 6 Day. h m 1 A.M. - 3 15 P.M. 3 30 2 | A.M. P. M. 4 15 ft. in, h m ft. in. 19 3.51.8 20" 8 11 18 4.7: 8 45 8 0.7 18 6 9 15 8 3 17 5.5 9 25 8 7 ft. in. 10 4.5 10 4 10 3 8 10-5 N.W. W.W. by N. N. by W. 5 2 “ The Time is Mean Time, obtained through the Royal Naval College; and the Dock-yard clock, which is used for ascertaining the l'imes of High and Low Water, is regulated thereby, by Mr. Smithers, Clock-maker, Portsca. “ The Height of the Tide is ascertained by Lloyd's Tide Gauge. “ The Line from which the Heights are measured is the Sill of the North Dock-gates."

·43 In the above specimen-register the force of the wind is indicated by figures, according to the following scale, contrived by Captain Beaufort: .0 Calm. 1 Light Air. Or just sufficient to give steerage way. 2 Light Breeze ... Or that in which a well-condi- 1 to 2 knots. tioned man-of-war with all 3 Gentle Breeze sail set, and clean full, 3 to 4 knots. would go in smooth water 4 Moderate Breczc.. from. 5 to 6 knots. 5 y'resh Breeze 6 Strong Breeze...... 7 Moderate Gale .... Royals, &c. Single-reefed topsails and top-gallant sails. Or that to which such a ship Double-reefed topsails, could just carry in chase, jih, &c. full and by Triple-reefed topsails, &c. Close-reefed topsails, and courses. ................ 8 Fresh Gale ....DOI. 9 Strong Gale......... 10 Whole Gale........ Or that with which she could scarcely bear close-reefed main topsail and reefed foresail. 11 Storin Or that which reduces her to storm stay-sails. 12 Hurricane Or that which no canvas could withstand. The circumstances of high water admit generally of more accurate observation than those of low water. The following table of the semimenstrual inequality serves to re-' duce observations made at any period of the lunation. Moon's Age. d 151 .......... h h h h Time of Moon's transit h Ol] h 3 2 5 hhhh h 6 7 8 9 10 11 .......... . m Lunitidal interval greater than mean (i. e. than corrected establishment)... 11-5-02-33--37-43-19 15,36 43 37 24 m Lunitidal interval less than at new or full moon (i. e. than vulgar establishment) 0 161 311 44 481 54 30 425|32|2613 less. greater. The determining the period of the lunation by the moon's age is a very inaccurate method; the mcthod by means of the moon's transit is much more exact. By the above table, the establishment of the place, either the cor- ·44 rected or the vulgar establishment, may be determined by tide observations made at any period of the lunation. But the above table is not accurate, nor can any table be accurate for all places: not only because the amount of the semimenstrual inequality is different at different places, (as has already been noticed,) but also because the epoch of the inequality is (slightly) different for different places. And these differences can only be determined by continued observations at cach place. But it is probable that the epoch of the inequality is very nearly the same in neighbouring places. . 4. METEOROLOGY. 1 The traveller should keep a complete meteorological register, skeleton forms for which have been prepared*. These are adapted for intervals of observation of six hours throughout the twenty-four; and although hourly observations be made, as is undoubtedly to bc desired, yet the regular entry and reduction of the observations for the hours in the skeleton forms is nevertheless essential, for the sake of future comparison with those similarly entered and reduced at the fixed stations. So many references havc lately been made to the Committee upon the subject of directions for meteorological observations, that they have embraced the opportunity of proposing a plan of extensive co-operation. In the way of general remark on this subject it may be observed, that it is impossible to pay too much attention to the zero points of the instruments, especially the barometer. The thermometers and barometers should be carefully compared with those of the Royal Society; and one barometer should be continually referred to as a standard, whenever the instruments are landed by the navigator, and on his return on board, so as to detect and take account of any change which may have occurred in the interval. Nor should an opportunity be lost, if it presents itself, of comparing the standard barometer (by the intervention of portable ones) with the standard barometer of the Cape Observatory, and with that used at Port Arthur, Van Diemen's Land, in the meteorological register kept by Sir J. Franklin's orders by Mr. Lempriere, as well as with the standard. at the observatory at Paramatta, and with any other instrument likely to be referred to as a standard or employed in research elsewhere. The general fact that the barometer at the level of the sea does not indicate a mean' atmospheric pressure of equal amount in all parts of the earth,-but, on the contrary, that the equatorial pressure is uniformly less in its mcan amount than that at and beyond the tropics,- was first noticed by Von Humboldt, and has since been de- Messrs. Richard and J. E. Taylor, (Red Lion Court, Fleet Street,) are directed to supply these skeleton forms at the charge of twenty shillings for 25 Pairs.

·45 monstrated by the assemblage of many observations made during voyages and on land by Schouw, as well as by other observations, an account of which will be found in the Reports of the Meteoro logical Committee of the South African Philosophical Society for '1836 and 1837. This inequality of mean pressure is a meteorological phenomenon of the greatest and most universal influence, as it is, in fact, no other than a direct measure of the moving force, by which the great currents of the trade-winds are produced; so that the measure of its amount, and the laws of its geographical distribution, lie at the root of the theory of these winds. The progress of barometric depression on approaching the line, and re-ascension in receding from it, will therefore be watched with interest proportionate to its intrinsic importance during the voyage outwards and homewards. But it may very well happen that phenomena purely local, of the same nature, may exist, not as cause but as effect; in other words, that the regular currents once established may, in particular localities, determined by the configuration of continents and by the influence of oceanic currents, or other causes, form permanent eddies or atmospheric ripples, so to speak, under which the mean pressure may deviate materially from the general average. An instance of permanent barometric depression of this kind, in the neighbourhood of the sea of Ochotzk, has been supposed by Erman; and a second seems to be pointed out in the neighbourhood of Cape Horn, by some remarks stated to have originated with Captain Foster; and it is not impossible that something of the same kind, but of an inverse character, may be found to obtain in that remarkable district. of Siberia mentioned by Erman, where, during winter, clouds are unknown and snow never falls; and it is somewhat curious to notice that the localities in question are not far from antipoles to cach other. In the passages of navigators across the equator (-specially should the ships be delayed by calms), opportunity may be presented of determining the amount of diurnal barometric fluctuation, apart from the interfering influence of land and sea breezes, or their equivalents far inland, which in all land observations encumber and disturb this somewhat obscure phenomenon; as well as for ascertaining, also apart from those influences, the existence or non-existence of that difference between the diurnal and nocturnal maxima and minima, which has been proved to exist in some localities, and sarmised to be general*. Connected with the equatorial barometric depression, and the ascensional current of heated air which produces that depression, is a phenomenon which may serve to elucidate the mechanism of this current in its origin, as well as to illustrate the mode in which ascending currents occasionally produce rain. It cannot be supposed that the whole body of the equatorial atmosphere rises en masse, or with any regularity or steadiness. Such a movement would be out of analogy with what we know of the movements of fluids in general. * See Reports of Met. Com. S. A. Phil. Soc. above referred to.

·46 Its tendency to rise is general, but this tendency is diverted by a thousand local influences, and concentred on particular points, where ît results in ascending columns and sheets, between which wind. flaws, capricious in their direction and intensity, and often amounting to sharp squalls, mark out the course of their feeders and of the indraft of cooler air from a distance to supply their void. Now the existence of such ascending columns is rendered frequently visible 'in a very unequivocal way, by vast piled-up masses of cloud of that peculiar form which has been called cumulostratus, the bottom being flat and ill-defined, the upper parts towering to an immense height, and ragged with great protuberances. From the bases of these great cumular piles are almost constantly seen to descend those vio. lent showers so common in the calm latitudes. It would be interesting on many accounts to obtain measures, even if somewhat vague, of the altitudes at which the bases of these clouds rest, as well as of the height of their summits, and to measure the temperature of the rain which falls from them at successive periods, as they pass over the ship, so as to ascertain whether the rain which falls along their axis be not colder (from coming at least in part from a greater elevation) than that from their skirts. The vapour plane, in such circumstances, being nearly or precisсly uniform over vast tracts of sca, the altitude of the base of such cloud vertically overhead may be considered the same as that of any other favourably situated for measure. In fact the determination of thc mean height of the vapour plane at and near the equator is one of high meteorological inport, and is connected by no circuitous steps with all the most interesting questions regarding the distribution of aqueous vapour over the globe and the irrigation of the continents. There can be little need to call the attentivn of navigators to anything relating to winds, storms, lightning, &c.; yet there are some points to which attention may bc expressly drawn, viz. to such distribution and movements of the clouds as indicate the existencc at the same time of an upper and an under current of wind moving in opposed or differing directions. In such cases, the sun, moon, or a star should be taken as a point to fix the eye. In storms the barometer should be very assiduously noted in relation to the varying phases of the gale and the changes of wind, and particularly to those sudden shifts of wind which characterize revolving storms. The Committee are not aware that the state of the barometer during a white squall" has ever been very carefully noted from instant to instant; or that it, or the more sensitive sympiesometer, have been' referred to during the approach and recess of a waterspout. The phenomena of ordinary thunder-storms may be thought to afford little matter for remark, and extraordinary ones will be noted of course. Yet there is one point to which we should wish that some attention might be paid, -it is the sudden gush of rain which is almost sure to succeed a violent detonation immediately over-head. Is this rain a cause or a consequence of the electric discharge ? Opinion would seem to lean to the latter side, or rather, we are not aware ·47 that the former has been maintained or even suggested. Yet it is very defensible. In the sudden agglomeration of many minute and feebly electrified globules into one rain drop, the quantity of electricity is increased in a greater proportion than the surface over which (ac- 'cording to the laws of electric distribution) it is spread. Its tension therefore is incrcased, and may attain the point when it is capable of separating from the drop to seek the surface of the cloud, or of the newly-formed descending body of rain, which, under such circumstances, and with respect to electricity of such a tension, may be regarded as a conducting medium. Arrived at this surface, the tension for the same reason becomes enormous, and a flash escapes. The following points should be observed, with a view to this mode of regarding the formation of lightning: 1st. The actual electric state of that rain which follows suddenly after a discharge originating vertically over head. 2nd. Whether lightning occurs without rain in the immediate point where it originates, or at least without a rapid formation and increase of cloud at that point. 3rd. Whether lightning proceeds from a cloud undergoing actual diminution from evaporation. 4th. Whether the cumular clouds, already noticed as continually forining and raining in the calm latitudes, send forth flashes of lightning; and if so, under what conditions, and with what effects. Observations of Auroræ form a highly interesting subject. Their effects on the magnetic needle should of course be narrowly watched; and all their phenomena should be minutely registered, such as the formation, colour, extent, situation, movement, and disappearance of arches, patches, banks, and streamers. In particular, attention should be drawn to an appearance which sometimes occurs, and which cannot but be regarded as highly instructive. It consists in pulsations propagated with more or less swiftness througi. patches of sky of definite forms, which liowever become visible only in successive portions, as thc pulse traverses thcm, giving the idea of masses of vaporous matter not visible per se, but rendered fitfully so, either by a Land of light cast in succession over every part of them from without, or by a temporary phosphorescence developed within their substance when traversed by electric matter. Such pulsations as above described formed very remarkable features of the auroras of October 12, 1833, and of January 18, 1839. Any indication of the near vicinity of auroral phenomena, or of their existence at a level below that of ordinary clouds, should be most minutely investigated at the moment, and carefully and circumstantially recorded. On the nights from the 11th to the 14th of November the sky should be watched for the periodical ineteors, whose existence seems now to be placed out of doubt; as also from the 9th to the 13th of August; and in general any remarkable display of shooting stars should be noted. The zodiacal light also should be observed in clear nights, with a view to the better defining its limits, and ascertaining if it be really, as ronie have supposed, variable in its extent 0 ·48 or lustre. Remarkable halos, parhelia, and other atmospheric phenomena, should be recorded, and careful measures of their dimensions taken with sextants or other instruments. 5. DISTRIBUTION OF TEMPERATURE IN THE SEA AND LAND. . Connected as this subject is with meteorology, it requires in some points of view to be considered apart. As the currents in the atmosphere are produced by the difference of temperature in its polar and equatorial regions, so it may be contended are those of the ocean by differences of temperature due to the same geographical causes. Such is the view taken by M. Arago in his elaborate instructions for the voyage of the Bonile, and it would appear undoubtedly more just than that which attributes them wholly to the friction and pressure of the winds. Nevertheless it must not be forgotten that there is an essential difference in the modes of action in the two cases. The sun's heat is effective in increasing the temperature of the air mainly from below, where it is in contact with the carth or water, both of which absorb the rays and compunicate them to the air above. In the sea the case is otherwise : for the sun's rays are totally absorbed at the surface, and no ray reaches the bottom of any sca deserving the name. The temperature therefore of a deep stratum of water cannot be permanently maintained by the sun's direct heat greatly above what it would have independently of its direct action. Herre the motive power in a system of currents so originating must be sought, not in the ascensional force at the equator, but in a descensional onc in the polar regions, or rather in that one polar region in which winter prevails. Thic order of the phenomicilu then is precisely the reverse of what obtains in the atmosphere; moreover the seaof the efficient agency is not only much less extensive than in the case of the atmospheric currents, but also subject to a semiannual shifting from one to the other extremity of the carth's axis, both which causes must tend greatly to diminish the average energy of the effect. Practically speaking, the question resolves itself into one of fact, which observation only can decide. Is there in the whole column of watershetween the surface of the occan and its bed at the poles, as compared with a column of equal depth at the equator and in free communication with it, a desccnsional power or not? and what is its amount? These questions can only be resolved by observations of the temperature and saltness of the sea, at various and considerable depths, in different latitudes, and under a great varicty of local cir. cumstances. The procuring such observations, and the preservation of specimens of the water, or the determination on the spot of their specific gravities, will afford a useful occupation in calms, and may be recommended as well worthy of attention. Theoretically speaking the subject is more complicated than at first appears, since it cannot but be that some considerable portion of solar heat absorbed by ale ·49 equatorial continents,-in place of finding its way out of the carth by radiation at the poles, in the mode of subterraneous communication suggested by Fourier, -inust escape through the bed of the ocean into its waters, and so be carried into their circulation. The distribution of temperature over the globe must greatly depend on the intensity which the solar rays possess on attaining the surface of the earth after traversing the atmosphere in different latitudes. TS subject this point to direct inquiry in a mode which after many years' trial has been found to give very satisfactory results, Actinometers are provided, and accompanied with very precisc directions for their inst". They should be observed only when the sky in the immediate neixhbourhood of the sun is perfectly free from visible cloud. On the other hand, depression of temperature caused by diurnal and nocturnal Biation by tlic only means we at present possess for that purpose, "... that of thermometers blackeneil and exposed in reflectors to the k;. --Will form a useful and valuable supplement to the actinometric resurcles. With a view to the collection of facts illustrative of the distribution of tenperature on land, wherever a ship may touch with a spect of reinaining some days, no time should be lost, on lunding, in burying in the earth one or inpre bottles (filled with spiritx, if there should be danger of water freezing), well packed in Sine's, or boxes stuffed with non-conducting inatter, such as woollen cloth, pounded charcoal, &c., but so as to leave easy access to the nich, which should be wide enough to introduce the bulb and stem of a good therinometer, so as to take the temperature of the contineri liquid rapidly, before it can have become altered by exposure Hiir (s taking up the bottle. Bottles so arranged should be Uuriul at depths of three, six, nine, twelve, &c. feet*, according to the facilities of penetrating the soil, and allowed to remain till the Yhdit of departure, so is to ensure their acquiring un precise temperaturo vf the soil; and whicn taken up should inmediately have the lemperature or the included liquiils ascertained. In case of very prolonged sojourn, monthly rendings should be taken. The tem- Peratura's of all springs and wells should also be diligently noted and registered. Opportunities for deternining the te. perature of the ocean at creat depths must of course be rare; but at moderate depths it can always be dune with comparatively little arouble, and the Committoe therefore, suggest the propriety of making observations of this Clement at two moderate and constant depthis (say 150 and 300 fathrms), by the aid of a self-registering thermordeter attached to a sounding line whenever the ship's way shall be such as to allow their being made with precision. 6. CURRENTS OF THE OCEAN. These are either subuqueous or superficial, and, like those in the atmosphere, both may coexist at the same place, with different direc- * These are the deptlis adopted in Mr. Forbes's recent experinients.

·50 tions and velocities. Öf the former we know almost nothing, and of the latter but little compared with what would be desirable and most useful. The practice of daily throwing overboard a bottle corked and sealed with the latitude and longitude of the ship at noon ought not to be neglected. A single instance of such a record being found may suffice to afford indications of the utmost value, while the trouble and cost are too trifling to mention. As no sea can be supposed absolutely motionless, the presence of a shoal, by casting up at the surface water which, but for it, would have continued to sweep along at a greatly lower level with the general body of the current, must bring the temperature of the surface water into nearer correspondence with that below. In low latitudes the surface water is hotter than that below; and accordingly it is a general remark, that the temperature sinks as the water shoals, or even in passing over banks whose depth is very considerable. If this theory of the phenomenon be correct, the contrary ought to be observed in situations where the surface water is colder than that below, as it is known to be under particular circumstances in the Polar Seas. In still Jarger tracts in high latitudes the seas have nearly a uniform temperature throughout their whole depth. In such circumstances should any superficial variation of temperature be observed in passing over a shoal or bank, it could only be ascribed to radiation. The subject is one of considerable interest to the navi. gator, as the approach to land or to shoal water is indicated by the thermometer with a high degrec of sensibility. Recent observations of this kind exist, the one at entering Table Bay in 1834, the other at quitting it. In the former casc the temperature fell 9° Fahr. in passing from deep water into the Bay; in the latter under reverse circumstances a rise of no less than 13° Fahr. was experienced, the temperature of the air remaining unaltered. The last-mentioned observations being very remarkable, the particulars are annexed. Memorandum of Observations made on board the Earl of Hardwicke, H.E.I.C.S. by Captain Henning. March 17th, 1839. Temperature of air at 51 1'.M., four miles from Cape Town.. 640.0 Of Sra. 520.0 5h 30m One mile north of Robben Island. Air 640.0 570.0 March 18th at sea. Air 64° 0 Sea 650.0 Sea .. Connected with the transcalescence of the air, is the transparency of the sea. The stimulus of the solar light no doubt affects the surface of mollusca at great depths, and numerous points of physical inquiry would be elucidated if we knew the co-efficients of extinction of the solar rays by pure sea water. As far as the luminous rays are concerned (or at least the chemical), the actual intensity of these rays at various depths might be very easily ascertained, both ·51 for direct sunshine and that of cloudy daylight, by the aid of Mr. Talbot's sensitive paper; which, duly guarded from wet by varnish and interposition between glass plates, might be sunk, face upwards in a small frame, while a portion of the same paper, cut from the same sheet, should be similarly exposed on deck, and partially shaded, inch by inch, from minute to minute, (or for a smaller interval according to the sensitiveness of the paper, with a view to inmediate com: parison between decks, by a light not strong enough to alter the tint. A mode of photometric measurement may be also furnished by the sensitive paper above alluded to, by the exposure of a small portion of it to the sun at noon for a given timne, suppose ten seconds, and subsequent comparison with a scale of tints. Paper duly prepared for these purposes was supplied for the use of the Antarctic expedition. During solar eclipses such paper ought to be exposed at intervals of five ininutes. The temperature of the soil under the direct, influence of the sun as indicated by a thermometer barely covered with dry earth, is an element of importance to the botanist, and may be recommended as an apt accompaniment to actinometric observations, The thermometer used should have a scale reading at least to 180° Fahr. The height of the line of perpetual snow, by whatever indications marked, should also be ascertained, wherever practicable. 7. DEPTH OF THE SEA. Soundings to as great a depth as practicable should be taken wherever opportunities may offer. Great difficulty, however, is well known to exist in the way of procuring any exact result, or indeed any result at all in very deep seas: and various methods (all objectionable) have been proposed and tried. The maximum depth of the sca is a gcological datum of such value, that a few failures incurred in attempts may very well be tolerated when placed in competition with the interest of even partial success. 8. VARIABLE STARS. In clear weather many interesting observations might be made by any one acquainted with the constellations, prprovided with a celestial map, as to the comparative lustre and variability of stars. Especially we would point out to the attention of such an observer the stars a Hydræ et Crateris as certainly, and a Orionis as probably, variable; the former at its greatest brightness being equal to e Argûs, and at its Icast equal or somewhat inferior to 8 Argûs, which are the best stars for comparison with it. Its period of change, however, being only very imperfectly known, additional observations would be valuable. The reinarkable star n Argûs should also be compared with others of the same apparent brightness, or nearly so, with a view to continue the history of its late extraordinary change of lustre. And the Committee carnestly recommend to any one who may undertake such observations, to form a list of a certain moderate number of stars, graduating 2 ·52 from the first magnitude downwards by almost insensible steps, and having rendered himself familiar with them, to note their arrangement in order of brightness,—not once only, but on a great many nights, forming on each occasion separate independent judgements ; trusting on no account to any printed catalogue, and diligently recording and prescrving his memoranda. Such observations are not fjart of the ordinary business of astronomical observatories, and are therefore neglected and abandoned to the amateur, the traveller, or the seaman in his night-watches, which they will be found to beguile of much of their tedium, and to reward by the frequent detection of variable stars not previously recognised as such. 9. REFRACTION. The determination of refractions nicar the horizon, both of celestial and terrestrial objects in high latitudes, is a very interesting subject of study. They may be pursued in various modes, of which perhaps the easiest is to note the disappearance of particular stars behind the horizontal edge of a board, erected at some considerable distance from a fixed point of observation, and then to ascertain, with all precision, the altitude of the line of disappearance, accompanying such observation with the height of the barometer and thermometer. Vertical diameters of the sun or inoon, when very near the horizon, with the corresponding altitudes, will also be of use, as well as measurements of the distances of two considerable stars on the same vertical, and direct measures of the altitudes of one and the same star in the progress of its diurnal course when near the horizon. The curve of terrestrial refraction might also be actually traced out by a levelling-tult. Any cases of me usual refraction, mirage, reduplication aut inversion of images, and of lateral refraction, should be recorded. 10. ECLIPSES. In annular or total solar eclipses the optical circumstances attending the formation and rupture of the ring should be minutely attended to, as well as the defalcation of light and heat, to be measured by their appropriate methods, as detailed in the Meteorological instructions. In lunar total eclipses the occultations of stars, whether large or small, should be looked for, and any apparent projection on the disk noticed. Great attention should also be paid to the intensity, colour, and distribution of illumination over the disk during the totul eclipse, as indicative of the gencral state of the earth's atinosphere in that great circle of the globe which at the moment is at right angles to the visual ray.

·53 INSTRUCTIONS FOR MAKING METEOROLOGI- CAL OBSERVATIONS. Arrer maturely considering this subject, the Comınittec do not presume to anticipate that what they may suggest will not be liable to objections, for their object will be to include within their compass many excellent series of observations which are already in progress, rather than to propose a degree of theoretical perfection, the attainment of which the present state of the science may not perhaps adinit of. Systematic co-operation is the essential point to which at present everything else should be sacrificed ; and co-operation on almost any plan would most certainly be followed by more beneficial results than any number of independent observations, however perfect they might be in themselves. The plan of co-operation should, in fact, be regarded at present as merely temporary and preparatory; but if steadily adhered to for a few ycars, it would certainly furnish the most perfect data for its own correction, which could then from time do time be applied with facility and precision. The Conmittec are not without liopes that amateurs of science may be induced to conform to these suggestions, even at the temporary sacrifice of their own views and convenience; for no one can reflect on the immense aniount of labour which is now rendered uscless for want of the requisite uniforinity and precision, without being convinced of the necessity of remedying an evil which has already been of too long standing, and continues to be a reproach to science. Many, of course, will not have it in their power to fill up the plan in all its details; but they will contribute greatly to forward the design, if, in such observations as they may find it convenient to make, they strictly comply with the rules proposed. They will be further encouraged to lend their aid to a comprehensive system, by the consideration that it will be adopted by the Government Observatories, as well as by those about to be established by the East India Company, and will of course be acted upon in the comparison and discussion of the observations inade at these institutions by the scientific authorities who will be entrusted with the execution of this task. The suggestions which the Committee wish to offer will relate, Ist, to the times of observation; 2ndly, to the situation of the instruinents to be observed; 3rdly, to the correction of the observations; 4thly, to a form of registry, which inay place many of the results in a striking point of view, and facilitate comparisons. 1. BAROMETERS. Times of observation. The purposes of meteorological observations would be most perfectly and most expeditiously obtained by hourly observations throughout the year; but since in the case of private observers in general, and in few public establishments, such ·54 a course of unremitting'labour cannot be hoped for, it is necessary, for general purposes, to select periods at longer intervals calculated to embrace the extremes of the periodical oscillations to which the pressure of the atmosphere is subject, and to ensure that uniformity of system at different stations on which the value of such observations so much depends. It is probable that the hours of 3. A.M., 9 A.M., 3 P.M., and 9 P.M., nearly coincide with the daily maxima and minima of the barometric column at the level of the sea, over a large portion of the globe; and it is desirable that as extensive a coinparison as possible should be instituted at these hours. At the magnetic observatories, it is provided that obscrvations shall be made every second, or even, hour of Gottingen mean time throughout the twenty-four; so that there at least, and in all others which will act in concert and correspondence with then, the complete diurnal cycle will be satisfactorily observed. It would be uselessly super-adding labour to the already extensive task imposed on these establishments, to require observations also at the hours above recommended for general adoption as Meteorological hours. They will, therefore, content themselves with filling up the forms furnished them, as adapted to the meteorological hours, with observations inade at the nearest magnetic hours to those named at each station. It is not, however, too much to expect that hourly observations should be made, during 24 hours, once in every month, by those who profess to pursue meteorology in a scientific manner; and when this cannot be effected, it is of the utmost importance that they should be made at least four times in the year, namely, at the sunimer and inter solstices, and at the spring and autumu equinoxes. One of the results of these hourly observations would probably be the indication of the exact times of the daily maxinia and minima of pressure at different stations, which, if :ut luund to coincide with the hours provisionally adopted, Diight ultimately be substituted for them ainder future directions. At the magnetic observatories the instruments will be read off' hourly, on the days set apart in each month for the Magnetic Termobservations, and the two-hourly system of observation in all cases continuing uninterrupted, will in effect furnish corresponding observations on all other days, whether arbitrarily chosen to suit private convenience, or in pursuance of the system about to be proposed in the subsequent paragraphs. Hourly sbservations at the equinoxes and solstices have been already instituted at numerous points both of Europe and America, at the suggestion of Sir John Herschel, whose directions should be strictly attended tå. They are as follows:- The days fixed upon for these observations are the 21st of March, the 21st of June, the 21st of September, and the 21st of December, being those, or immediately adjoining to those, of the equinoxes and solstices in which the solar influence is cither stationary or in a state of most rapid variation. But should any one of those 21st days fall on Sunday, then it will be understood that the observations are to be deferred till the next day, the 22nd. The observation at each station should commence at 6 o'clock A.M. of the appointed days, and ter- ·55 minate at 6 A.m. of the days following, according to the usual reckoning of time at the place. The commencement of cach hour should be chosen, and every such scries of observations accompanied by a notice of the means used to obtain the time, and, when practicable, by some observation of an astronomical nature by which the time can be ascertained within a minute or two. Travellers provided with meteorological instruments who may be stationary on any of these days, may use them with advantage on such opportunities. Such as may ascend high mountains are recommended, cæteris paribus, to choose one of these days as affording a greater probability of securing a complete series of corresponding observations than any other; for which reason these observations cannot be too strongly recommended to residents in mountainous countries. The geologist, nay even the surveyor, may find his account in traversing his field, barometer in hand, on one of these days, provided he have reason to presume that there exist observers in its neighbourhood who take a part in these observations. It is to be hoped that to scientific meteorological observers the six-hourly observations may not be found to be impracticable throughout the year; but in any case where it may be impossible to observe regularly at 3 A.M., an effort should be made to include that hour on the days of the new and full moon, and quadratures, or at least on the days of the new and full moon ;-as it must be borne in mind, that in what concerns the great meteorological questions on which the inost interesting features of the subject depend, the night is quite as important as the day, and has been hitherto far too niuch neglected. Whatever hours, however, may be selected for the regular series of observations, the greatest card should be take not to insert in the register anything deduced by interpolation from observations made at other hours, or anything in short but what has been actually observed. It is much to be wished that occasional observations niay be made under remarkable circumstances, such as during great rises or great falls of the barometer, at the period of great storms, earthquakes, &c.; but such obscrvations should be registered apart. The barometer should be placed in an apartinent subject to as little variation of temperature as possible, and in a good light; and to facilitate night observations, an arrangement should be made for placing behind it a light screened by a sheet of white paper, or other diaphanous substance. Great care should be taken to fix it in a perpendicular position by the plumb-line. Its height must be carefully ascertained above some permanent and easily-recoverable mark, either in the building in which it is situated, or in some more permanent building, or rock, in its immediate vicinity; and no pains should be spared to ascertain the relation which such mark may bear to the level of high and of low water at spring tides, and ultimately to the mean level of the sea. Changes in the adjustments of meteorological instruments should ·56 any 0 be most carefully avoided; but whenever alterations may be absolutely necessary, they should be made with all deliberation, scrupulously noticed in the register, and the exact amount of the change thence arising in the reading of the instrument under re-adjustment ascertained. As far as possible, registers of meteorological observations should be complete; but if, by unavoidable circunstances of absence, or from other causes, blanks occur, no attempts to fill them up by general recollection, or by the apparent course of the numbers before and after, should ever be made. The observatories established by her Majesty's Government are furnished with two barometers cach, of Newman's constructionthe one a standard, and the other portable; and they are accoinpanied by accurate directions for fixing and observing them. The standard instrument is of large dimensions, its tube being of the diameter of 0:6 inch. It requires two adjustments: 1st, The whole scale, which is of brass, is inoveable, and terminates in an ivory point, which is carefully brought down to the surface of the mercury in the cistern, and the two are known to be accurately in contact when the actual point and its reflexion appear just to touch one another. The scale is laid off from this point from an authentic standard. 2ud. The second adjustment is that of the vernier, in which the upper part of the scale terminates, to the surface of the incrcury in the tube. For this, both the back and front edge are made to coincide, and brought down so as to torni a tangent to the curve, and just to exclude the light between them at the point of contact. In making both these adjustments, it is desirable that the eye should be assisted by a magnifying glass. Before the observation is made the instruinent should be slightly tapped, to free the mercury fronı any adhesion to the glass; but any viulvat uscillation should be avoided. The portable barometer has only one adjustment, namely, that of the veruier to the upper surface of the mercury in the tube, which adjustment must be effected with the same precaution as in the casc of the standard instrument. This first reading may be entered in the column prepared for it in the register, and beside it the temperature of the inercury carcfully read off from the thermometer which dips into the cistern. In the case of the standard barometer the first measure is taken immediately from the surface of the mercury in the cistern, and it requires no correction for the different capacities of the tube and cistern. Neither docs it require any correction for capillary action, as the large diameter of the tube renders this correction inappreciable. The portable barometer, however, requires corrections for both these circumstances. For the purpose of the former, the neutral point is marked upon cach instrument, or that particular height which, in the construction of the instrument, has been actually measured from the surface of the mercury in the cistern. It is obvious that, in almost every case, the mercury will stand either above or below the neutral point: if above, a portion of the mercury ·57 must have left the cistern to enter the tube, and consequently must have lowered the surface in the cistern; if below, a quantity of mer. cury niust have left the tube, and, entering the cistern, raised the level of the micrcury in it. For the correction of observations for this circumstance, the relation of the capacities of the tube and cistern have been experimentally ascertained, and are marked upon the instrument: thus, capacity z'rth, indicates that for every inch of elcvation of the mercury in the tube, that in the cistern will be depressed one 50th of an inch. Thus, when the mercury in the tube is above the neutral point, the difference between it and the neutral point is to be divided by the capacity, and the quotient being added to the observed height, the result will be the corrected height. Or if the mercury at the time of observation should be below the neutral point, the difference of the two is to be divided as before, and the quotient to be subtracted from the observed height. Thus, suppose the capacity to be 'oth, the ncutral point 30 inches, and the observed height 30-500 inches, the difference is 0:5 inch, which divided by 50 gives 0:01 inch to be added to the observed height, producing 30:51, the corrected height; or if the observed height be 29 inches, the difference, 1 inch, divided by 50, gives •02 inch to be subtracted from the observed height, giving 28.980 inches for the corrected height. The second correction required is for the capillary action of the tube, the effect of which is constantly to depress the mercury in the tube by a certain quantity inversely proportioned to the diameter of the tube. In the instruments furnished to the fixed observatories the amount has been experimentally determined during the construction, and marked upon the instrument; the quantity is always. to be added to the height of the mercurial colunin, previously corrected as before. For the convenience of those who may have barometers, the capillary action of which has not been so determined, a table of the corrections for tubes of different diameters is placed in the Appendix, page 81. The marine barometers furnished to the Antarctic Expedition differed in nothing from the other portable barometers but in the mode of their suspension, and the necessary contraction of the tubes to prevent oscillation from the motion of the ship, and require the same corrections. When these two corrections have been Inatle in the first reading of the portable barometer, it should agree with the direct observation of the standard barometer; and it is very desirable that frequent comparative observations should be made of the two instruments, in order to ascertain whether thcre may be any permanent difference between then. Should this be the case, the amount may be marked upon the instrument, and allowed for as an index error, in order that, if an accident should happen to one, the other may be substituted for it without detriment to the regular series of observations. It is to be presumed that the portable barometer will frequently be employed in ascertaining the altitude of remarkable points in the vicinity of observatories. The instruments furnished to the observatorics established by her ·58 Majesty's Governmerft and the similar observatories in India have been all independently graduated and compared with the standard of the Royal Society; and in all cases it is desirable that such a comparison should be made with some standard instrument of authority, directly, or by means of a good portable barometer. In making such comparisons, all that is necessary is to record five or ten simultaneous readings of both instruments, deliberately made, at intervals of a few minutes from each other, after at least an hour's quiet exposure, side by side, that they may have the same temperature. If compared by two observers, each should read off his own barometer in his usual manner, then cach should verify the other's result. By this means the zero of one standard may be transported over all the world, and that of others compared with it ascertained. To do so, however, with perfect effect requires the utmost care in the transport of the intermediate barometer, and is by no means an operation either of trifling import or of hurried or negligent performance: some of the greatest questions in meteorology depend on its duc execution. The next correction, and in some respects the most important of all, is that due to the temperature of the inercury in the barometer tube at the time of observation. To obtain this every barometer requires to have attached to it a thermometer, which in the instruments furnished to the observatories dips into the mercury in the cistern, and this must be read and registered at each observation of the barometer. In the appendix pages 82, &c. will be found a table calculated by Professor Schumacher, which gives for every degree of the thermometer and every half inch of the barometer, the proper quanthuy to be added or subtracted for the reduction of the observed height to the standard temperature of the mercury at 32° Fahr. It must, however, be observeri, that this table is only calculated for barometers whose scales are engraven upon a rod or plate of brass reaching from the level of the mercury to the vernier. In many barometers the scale is engraved upon a short plate of brass fixed upon the wooden frame of the instrument, and the compound expansion of the two substances can only be guessed at, but must be obviously less than if the whole length had been of brass. As a near approximation for such inperfect instruments, another table has been placed in the appendix page 88, in which the lesser expansion of glass has been substituted for that of brass. No scientific ob. server, However, would Willingly use such an instrument. Although all these corrections are necessary for the strict reduction of registered observations, they ouglıt not to be applied to individual observations previously to registry. In the blank forms of register furnished to the observatories, one sheet is devoted to uncorrected observations, and a second to the corrected, and it is much to be wished that the proper reductions should be made as soon after the observations as possible. 2. THERMOMETERS. Times of observation. The external standard thermometer should be observed and registered at the same times as the barometer, and ·59 case. all the register thermometers may be read off at the time of the 9 A.M. observation, and their indices re-adjusted. But as double maxinia frequently, and double minima occasionally, occur, in consequence of sudden changes of temperature, both the thermometers should be occasionally inspected with a view to ascertain whether the motion of either the mercury or the spirit has been reversed in an unusual manner; and such double maxima or minima should be recorded apart as supernumerary, with the dates and leading features of the Each obscrvatory was furnished with a standard mercurial thermometer, which was carefully compared with an authentic standard, to be deposited at the apartments of the Royal Society. The spirit therinometers have been constructed with every possible attention to accuracy. A fixed point of temperature has been directly determined for each, at or about 0", by incans of a freezing iniature and comparison with a standard mercurial thermometer; and to render their indications accurate at still lower temperatures, the tubes are frustra of inverted cones, the rate of redaction of calibre being such as to compensate very nearly for the diminishing rate of the contraction of alcohol as we descend in the scale of temperature. For example, a column of Mercury 5-10 inches long at the top of the instrument, changes to 5:40-543-5.16 : the rapid change of the calibre at the top of the tube corresponding to the increased expansion of the spirits at the top of the scale. The specific gravity of the spirit employed is 822. In general, however, accuracy would be most satisfactorily insured by selecting tubes of equal borcs, and dividing the scales into equal degrees, which could be casily correcteil, for the diminishing rate of the contraction of alcohol, ly reference to proper tables. It is recommended that all thermometers be carefully and frequently compared with the standard, and their differences, at one or more temperatures (the wider asunder the better), marked upon their scales and applied as index errors. This is particularly necessary with the register thermometers, whose construction renders them inost liable to such errors. In placing the standard thermometer, an exposure should be chosen perfectly shaded from the sun, where no reflected sunbeams from water, buildings, rocks, or dry soil can reach it, and which is easily accessible for observation. It should be fixed, not merely hung, upon a bracket projecting six inches from the wall, or other support to which it may be attached, and it must be completely sheltered from rain by a screen, so that the bulb shall never be wetted. In reading it, the observer should avoid touching, breathing on, or in any way warming it by near approach of his person; and in night observations particular care should be taken not to heat it by approximation of the light. The quicker the reading is done the better. It is desirable that notice should be taken of all sudden and remarkable changes of temperature, although such occasional observations must not be recorded in the regular series.

·60 The self-registering thermometers should be placed with the same precautions as the standard, and so fastened as to allow of one end being detached, and lifted up to allow of the indices within the tubes sliding down to the ends of the fluid columns, which they will readily do with the assistance of occasional tapping. The self-registering thermometers are apt to get out of order by the indices becoming entangled, or from the breaking of the column of fluid. When this happens with the spirit thermometer, it may be rectified with case by jerking the index down to the junction of the bulb and tube. The whole of the tube will at the same time become wetted with the spirit; and by setting it on end with the bulb downwards, the spirit will run together into one continuous column. When the steel index of the mercurial thermometer becomes immersed in the mercury, it must be jerked in the opposite direction, till it, with the mercury which may be above it, is projected into the little bulb at the top of the tube. If this do not succeed, heat must be applied to the mercury-bulb, and when the index is fairly lodged in the air-bulb, by carefully warming the mercury-bulb with a spirit lamp naving a very small flame, the mercury inust he made to expand till it rises to the very top of the tube, and projects convexly into the air-bulb. The tube inust then be placed upright, and, by tapping, the detached inercury will slip down beneath the steel index, and will fairly unite with the convex projection aforesaid. Now let the bulb cool, and the mercury will sink in one united column, and leave the index free. Besides the regular series of observations of the temperature of the air, there are other occasional observations to be made of temperature under different circumstances, which might possess great interest. The surface temperaturi, of the water of the sea or of rivers may be conveniently obtained by taking up a bucket-full of water and stirring round the thermometer in it. The temperature of the water of deep wells may be ascertained in the same way, and should be taken monthly, if near the residence of the observer. The temperature of rain should also be attended to at times; it may be determined by receiving the rain in a linen funnel, totally cuclosed in a tin case to prevent cooling by evaporation from the linen. The temperature of the soil at different depths is a point of considerable importance. For this purpose excavations should be inade in a dry sheltered situation, 3, 6, and 9 feet deep, and lined with brick or carthenware tubes. In the bottom of these excavations carthenware quart bottles may be carefully placer, filled with water, spirit, or brine, and corked. They must be carefully covered with tow or cotton, and drawn up on the days of horary observation*, and their temperatures taken by an accurate thermometer, and registered apart. As a general caution it may be mentioned, that the standard ther- * See p. 38.

·61 mometer should never be exposed to risk by application to such purposes, but thermometers which have been compared and com rected by comparison with it. 3. ACTINOMETERS. Amongst the observations of highest importance must be ranked. those of the force of solar and terrestrial radiation. The most perfect means of observing the former is afforded by the actinometer. This instrument consists of a large hollow cylinder of glass, soldered at one end to a thermometer-tube, terminated at the lipper end by a ball clrawn out to a point, and broken off, so as to leave the end open. The other end of the cylinder is closed by a silver or silverplated cap, cemented on it, and furnished with a screw, also of silver, passing through a collar of waxed leather, which is pressed into forcible contact with its thread, by u tightening-screw of large diameter enclosing it, and working into the silver car, and driven home by the aid of a strong steel key or wrench, which accompanies the instrument. The cylinder is filled with a deep blue liquid (ammonio-sulphate. of copper), which ought to have been prepared some months beforehand, as it deposits a sediment when fresh, however clear or carefully filtered. This sediment, if deposited in the interior of the instrument, may be washed out with weak muriatic acid, which should itself be removed by water before refilling the instrument, and the ball at the top being purposely left full of air, and the point closed with melted wax, it becomes, in any given position of the screw, a thermometer of great delicacy, capable of being read off' on a divided scale attached. The cylinder is enclosed in it chamber blackened on three sides, and on the fourth, or face, detended from currents of air by a thick glass, renovcable at pleasure. The action of the screw is to diminish or increase at pleasure the capacity of the hollow of the cylinder, and thus to drive, if necessary, a portion of the liquid up into the ball, which acts as a reservoir, or, if necessary, to draw back from the reservoir such a quantity as shall just fill it, leaving no bubble of air in the cylinder. To use the instrument, examine first wlicther there be any air in the cylinder, which is easily seen by holding it level, and tilting it, when the air, if any, will be seen to run along it. If there be any, hold it upright in the left hand, and the air will ascend to the root of the thermometer-tube. Then, by alternately screwing and unscrewing the screw with the right land, as the case may require, it will always be practicable to drive the air out of the cylinder into the ball, and suck down liquid, if any, from the ball, to supply its place', till the air is entirely evacuated from the cylinder, and the latter, as well as the whole stem of the thermometer-tube, is full of the liquid in an unbroken column. Then, holding it horizontally, face upwards, slowly and cautiously unscrew the screw, till the liquid retreats to the zero of the scale. The upper bulb is drawn out into a fine tube, which is stopped with wax. When it is needed to empty, cleanse, and refill the instru- ·62 ment, liquid must first be forced up into the ball, so as to compress the air in it. On warming the end, the wax will be forced out, and the screw being then totally unscrewed, and the liquid poured out, the interior of the instruinent may be washed with water slightly acidulated, and the tube, ball, &c. cleansed, in the same way, after which the wax inust be replaced, and the instrument refilled. To make an observation with the actinometer, the observer must station himself in the sunshine, or in some sharply terminated shadow, so that without inconvenience, or materially altering his situation, or the exposure of the instrument in other respects, he can hold it at pleasure, cither in full sun or total shadow. If placed in the sun, he must provide himself with a screen of pasteboard or tin plate, large enough to shade the whole of the lower part or chainber of the instrument, which should be placed not less than two feet from the instrument, and should be removeable in an instant of time. The best station is a room with closed doors, before an open window, or under an opening in the roof into which the sun shines freely. Draughts of air should be prevented as much as possible. If the observations be made out of doors, shelter from gusts of wind, and freedoni from all penumbral slalows, as of ropes, rigging, branches, &c. should be sought. Generally, the more the obscrver is at his ease, with his watch and writing-table beside him, the better. He should have a watch or chronometer beating at least twice in a second, and provided with a second hand; also a pencil and paper ruled, according to thc form subjoined, for registering the observations. Let him then grasp the instrument in his left hand, or if he have a proper stand (which is preferable on shore or in a building *), otherwise firmly support it, so as to expose its face perpendicularly to the direct rays of the sun, as exactly as may be. The liquid, as soon as exposerl, will mount rapidly in the stern. It should be allowed to do so for three yo tour winutes before the observation begins, taking care, however, not to let it mount into the bulb, by a proper use of the screw. At the same time the tube should be carefully cleared (by the same action) of all small broken portions of liquid remaining in it, which should all be drawn down into the bulb. When all is ready for observation, draw the liquid down to zero of its scale, gently and steadily; place it on its stand, with its screen before it, and proceed as follows. Having previously ascertained how many times (suppose 20) the watch beats in five seconds, let the screen be withdrawn at ten seconds before a completo minnte shown by the watch, suppose at 2h 14m 50%. From 50% to 55%, say 0, 0, 0, .... at cach beat of the watch, looking meanwhile that all is right. At 5.56 complete, count 0, 1, 2, up to 20 beats, or to the whole minute, 215m 0%, kecping the eye not on the watch, but on the end of the rising column of liquid. At the 20th beat read off, and register the reading (12° 0), as in co- * This may consist of two deal hoards, 18 inches long, connected by a hinge, and kept at any required angle by an iron, pointed at each end. The upper should have a little rahbet or moulding fitting loosely round the actinometer, to prevent its slipping off.

·63 lumn 3, A, of the annexed form. Then wait, watching the column of air above the liquid, to see that no blebs of liquid are in it, or at the opening of the upper bulb (which will cause the movement of the ascending columu to be perforined by starts), till the minute is nearly elapsed. At the 50th second begin to watch the liquid rising ; at 55% begin to count 0, 1, 2, up to 20 beats, as before, attentively watching the rise of the liquid; and at the 20th beat, or complete minute (2h 16m 08) read off, and instantly shade the instrument, or withdraw it just out of the sun and penumbra. Then register the reading off (43°3) in colunin 3, B, and prepare for the shade observation. All this may be done without hurry in 20 seconds, with time also to withdraw the screw if the end of the column be inconveniently high in the scale, which is often required. At the 20th second prepare to observe; at the 25th begin to count beats, (), 1, 2, .... 20; and at the 20th beat, i. e. at 2" 16" 30%, read off, and enter the reading in colunm 3, A, as the initial shade reading (4592). Then wait as before till nearly a minute has elapsed, and at gh 17 20$ again prepare. At 17m 25% begin to count beats; at 17m 30s read off, and enter this terminal shade reading (420.8) in coluinn 3, B, and if needed, withdraw the zero. Again wait 20%, in which interval there is time for the entry, &c. At 17m 50$ remove the screen, or expose the instrument in the sun; at 55* begin to count beats; and at the complete minute, 180, read off' (1408), and so on for several alternations, taking cire to begin and end cach series with a sun observation. If the instrument be held in the land, care should be taken not to change the inclination of its axis to the horizon between the readings, or the compressibility of the liquid by its own weight will produce a very. appreciable ainount of error. In the annexed form column l. contains the times, initial and terminal of each sun and share observat.on. Colum: 2. (xpresses by 3. 4. -- 6. 1. Date and times of observation. Fel,'hausen, 1837, Oct. 30. --- Exposure, sunco) Readings of the Instrument. Change Radia. per tion in minute. parts of B - A. scale, Remarks. A. B. Initial. Terminal. Tuitial. 1 crminal. S S h in ni 2 15 0 16 0 16 30 17 30 18 0 19 (1) 19 30 20 30 210 22 0 22 301 23 30 210 25 0) OXOXOXO The times are reduced to appa- + 12.0 + 433 + 31:31 rent time, or to the sun's hour 45.2; 42.8 24 34.75 augle from the incridian, 14.8 48-2 + 334 33:40 (Zero withdrawn. 28.0 26.8 1.4) 34.85 43:9 + 33.5 34.75 46-6 45-5 1.1 34.95 9-0 43.2 + 342 General mean per formula = 3473 for 2h 2000s of apparent time. an appropriate mark, o and X, the exposure, whether in sun or shade. Column 3. contains the readings, initial and terminal (A and B). Column 4. gives the values of B - A, with its algebraical ·64 sign expressing the rise and fall per minute. And here it may be observed, that if by forgetfulness the exact minute be passed, the reading off may be made at the next 104, and in that case the entry in column 4 must be not the whole amount of B A, but only this of that amount, so as to reduce it to an interval of 60% procisc. Colunin 5. contains the radiations as derived from successive triplets, O XO, XOX, O XO, &c. by the formula presently to be stated; and in column 6. are entered remarks, such as the state of the sky, wind, &c.; as also (when taken) the sun's altitule, barometer, thermometer, and other readings, &c. The formula of reduction is as follows. Let O. X: On x', ", x", &c. represent the numbers in colunin 1, with their signs in order, as they stand, or the al pos of B A. Then will the numbers in column 5 be respectively, х +'- - *** +01.0"-x' + x' + x +0", 2 and so on, the algebraic sigus bwing carefully attended to. Thus 31.3 + 33:4 34-75 = + + 2 3.5.70 + 24 + 1.1 fins, d'o. 2 The mean of a series not excecding tliree or four triplets may bir had by the formula O + O' + 0" * &c. x + x' * &c. ገዜ 20 I where w is the number di sim observations, the time corresponding being the middle of the middle shade observatiou. A coniplete actinometer observation cannot consist of less than three sun and two shade observations interinciliate; but the more there are taken the better, and in a very clear sunny day it is highly desirable to continue the alternate observations for a long time, even from sunrise to sunset, so as to deduce by a graphical projection the law of diurnal increase and diminution of the solar radiation, which will thus readily becoine apparent, provided the perfect clearness of the sky continue,-an indispensable condition in these oliserrations, the slightest cloud or haze over the sun being at once marked by a diminution of resulting radiation.

·65 To detect such haze or cirrus, a brown glass applied before the eye is useful, and by the help of such a glass it may here be noticed that solar halos are very frequently to be seen when the glare of light is such as to allow nothing of the sort to be perceived by the unguarded eye. It is, as observed, essential that the instrument be exposed a few minutes to the sun, to raise its temperature in some slight degree. If this be not done, owing to some cause not very obvious, the first triplet of observations (sun, shade, sun) will give a radiation perceplibly in defect of the truth, as will become distinctly apparent on continuing the series. But it may be as well for a beginner to conimenee at once reading as soon as the instrument is exposed, and reject the first two triplets, by which he will see whether he has all his apparatus conveniently arranged, and get settled at his post. When a series is long continued in a good sun, the instrument grows very hot, and the rise of che liquid in the sun observation decreason, while the fall in the shade increases; nay, towards sunset it will fall even in the sun. This phenomenon (which is at first startling, and seeming to impeach the firlclity of the instrument) is, in fact, perfectly in order, and produces absolutely no irregularity in the resulting march of the radiation. Only it is necessary in casting up the result (in col. 5.) to attend carefully to the algebraic signs of the clifferences in column 4, as in the following example (which, as well as that abore given, is one of actual occurrence). 3. 5. 6., 1. 2. Date and times of ob servation, Wynberg, Nov.21, Exposure, 1837 shade. Readings of the Instruinent. sun or Change Radia. per tion in mitute. parts of B-A. scale. Remarks, 'I'ermal B, Initial. Terminal. lii ১ > 9 ()1 10 10 30 11 30 12 0 13 0 13 30 U :3!) () 16 0 6 19 15 : CxCxO: Alt. of O = 7° 19'. tot 90+ 9.7 + 0.7' 10-8 12 11:23 34.0 31.1 209.25 28:5 170) 11:3! 8.20 Cirrus haze co. 12:01 ming on. Alt. of = 1037. Every series of actiuometer observations should be accompanied with notices in the column of remarks of the state of the wind and sky gencrally, the approach of any cloud (as seen in the coloured glass) near to the sun; the barometer and thermometers, dry and wet, should especially be read off more than once during the series, if a long one, and, if kept up during several hours, hourly. The tipies should be correct to the nearest ninute, at least as serving to calculate the sun's altitude; but if this be taken to the nearest minute or two) with a pocket sextant, or even by a style and shadow, frequently (at intervals of an hour or less) when the sun is rising or F ·66 setting, it will add much to the immediate interest of the observations. When the sun is near the horizon, its reflection from the sca, or any neighbouring water, must be prevented from striking on the instrument; and similarly of snow in cold regions, or on great elevations in alpine countries. Every actinometer should be provided with a spare glass, and all the glasses should be marked with a diamond; and it should always be noted at the head of the column of remarks; which glass is used, as the co-efficient of reduction from the parts of the scale (which are arbitrary) to parts of the unit of radiation varies with the glass used. In the case of the actinometers sent out with the Expedition and to the fixed observatories, it was not practicable to ascertain these co-efficients for each instrument and cach glass, owing to the total absence of any favourable opportunity of sunshine. The values of the parts of the respective scales of the instruments, as determined approximatively by careful measurement of the dimensions, were as follows: Marks of the Actinometers. Multiplier for reducing parts observed to parts of Approximate vaa standard retainlue of one part of scale in Actines. ed in possession, marked A 1. Maker's Mark. Private Mark. 1 2 3 A 4 A 5 6 K L M N O 1 1.4909 1.3726 1.4020 1.6.550 1.4403 1.0608 7.085 6:523 6•663 7.864 6-841 5-041 The dimensions of the instruments which are used in these reductions are, 1st. The external diameter of the cylinder containing the coloured liquid, i. e. its mean diameter, if on measurement with fine callipers its two ends be found to differ. 2nd. The length of thał portion of it which receives the sunbeam. The product of these two data gives the area of the section of the sunbeam effective in raising the temperature, and which, though not all equally effectiye, by reason of the cylindrical form of the glass, is yet effective in the same ratio in all of them by reason of their general similarity of figure. 3rd. The content in water grains) of 100 parts in length of the capillary tube used for the scale. This may best be determined by gauging it with mercury before it is soldered to the cylinder, and ought always to be so determined by the maker; but when fitted, this is impracticable, and the measurement of the element in question must be performed as follows:

·67 The instrument being placed horizontally, and allowed to attain the precise temperature of the apartment, let the liquid be brought to zero by the motion of the screw; after which let the screw be turned precisely one revolution, or half revolution (as the scale may require) in, and note the rise of tle liquid in parts of the scale. This must be done several times, alternately screwing in and out. The screw must then be taken out; its threads counted, and the weight of water displaced from a narrow vessel cxactly full, by the immersion of the whole length occupied by the thread exactly ascertained by a nice balance; after which a very simple calculation will give the value of the parts of the scale in water grains required; this process was followed in the case of the instruments above mentioned, and if carefully conducted is susceptible of great precision. The glasses as well as the cylinders and capillary stems of the instruments, if accidentally broken, should have their fraginents carefully preserved and labelled. Thc unit of solar radiation to be adopted in the ultimate reduction of the actinometric observations is the actine, by which is understood that intensity of solar radiation, which at a vertical incidence, and supposing it wholly absorbed, would suffice to melt one millionth part of a metre in thickness, from the surface of a sheet of ice horizontally exposed to its action per minute of mean solar time; but it will be well to reserve the reduction of the radiations as expressed in parts of the scale to their values in terms of their unit until the final discussion of the observations. Meanwhile, no opportunities should be lost of comparing together the indications of different actinometers under similar and favourable circumstances, so as to establish a correspondence of scales, which in case of accident happening to one of the instruments, will preserve its registered observatious froin loss. The comparison of two actinometers may be executed by one observer using alternately each of the two instruments, thus, Instrument A. Instrument B. A. Etc. O...... O..... OXO x 0 O beginning and ending with the same; though it'would be more conveniently done by two observers observing simultaneously at the same place, and each registering his own instrument. An hour or two thus devoted to comparisons in a calm clear day, and under easy circumstances, will in all cases be extremely well bestowed. Neither should cach observer neglect to determine for himself the heat stopped by each of his glasses. This may be done also by alternating triplets of observation made with the glass on and off, thus, F 2 ·68 Glass off. Glass on. Glass off. Etc. O O. X OXO O beginning and ending with the glass off, and (as in all cases) begin. ing and ending cach triplet with a sun observation. For the purpose now in question a very calm day inust be chosen, and a great inany triplets must be taken in succession. It will be found that a single thickness of the ordinary blueish or greenish plate glass stops about 0-20 (= }) of the incident calorific rays; a second glass about 0-16 (or a materially less proportion) of those which have escaped the action of the first. No two glasses, however, are precisely alike in this respect. Very interesting observations may be made by two observers furnished with well-conipared actinometers, the one stationed at the summit, the other at the foot of some great elevation, especially if the stations can be so selected that the observers shall be nearly in the line of the incident sunbeam at the time of observation, so as both to lie in the atinospheric column traversed by the rays. Many convenient stations of this kind might be found in mountainous countries; and by repeating the observation two or three times under favourable circumstances, interchanging obscrvers and instruments, &c., and accompanying the observations with all circumstantial and local elements of precision, there is no doubt that the co-efficient of extinction of solar leat in traversing at least the lower strata of our atmosphere might be obtained with nuch exactness, and thus a highly valuable datum secured to science. The observers tould, of course, agree w make their observations strictly simultaneous, and should, therefore, compare watches before parting. The actinomctor is also well calculated for measuring the defalcation of heat during any considerable eclipse of the sun, and the Committee would point out this as an object worthy of attention; as many eclipses invisible or insignificant in one locality, are great, or even total thers. The observations should commence an hour at least before the eclipse begins, and be continued an hour beyond its termination, and the series should be uninterrupted, leaving to others to watch the phases of the eclipse. The atmos, pheric circumstances should be most carefully noted during the whole series. Though out of the question in the circumstances immediately under contemplation, it may not be amiss to remind aeronauts, that observations of the actinometer may, no doubt, be made with considerable case and precision in the car of a balloon, and if accompanied with good barometric and hygrometric simultaneous observations aloft and below, as well as with a careful and copious registery of the temperatures of the air corresponding to each successive step of depression in the barometric column, would in every ·69 point of view be most precious, especially as respects the intricate problem of astronomical refractions for the solution of which such a series of data could not fail to be most valuable, thus adding one to the many useful subjects of inquiry in those hitherto almost useless adventures. For the purpose of securing a steadily progressive ascent, it would be desirable that the ballast should be discharged continuously and not in large quantities at a time. 4. RADIATING THERMOMETERS. As the actinometer can only be observed at intervals in perfectly clear weather, additional information with regard to solar radiation, of much interest, though not of so precise a nature, may be obtained, by the daily register of the maximum temperature of a register thermometer, with a blackened bulb exposed to the full action of the sun's rays. It may be placed about an inch above the bare soil, and screened from currents of air. The maximum temperature indicate by such a thermometer, even in cloudy weather, will generally be considerably above that of the air, and the maxima and inean daily maxima of its indications will, after a long scrics of observations, afford data of the utmost value to the history of climates. The bully of the thermometer should be about half an inch in diameter, and it may be uniformly blackened with lampblack and varnish. The graduation should be made upon the glass steni, to prevent any inconvenience from the expansion and warping of the scale. The measure of terrestrial radiation is of no less importance to the science of meteorology than that of solar radiation, but no perfect instrument has yet been contrived for its determination. Very va: Juable information, however, may be derived from the daily register of the inininum temperature of a register spirit-thermometer, the bulb of which is placed in the focus of a parabolic metallic mirror, turned towards the clear aspect of the sky, and screened from currents. The mirrors furnished to the observatories are of silverplated copper, but planished tin-plate or zinc might be substituted without detriment. They are 6 inches diameter and 2 inches deep, and the thermometers which are graduated upon the stems pass through sockets in their sides, in which they may be accurately adjusted by corks. Their bulbs do not exceed half an inch in diameter. Even in the daytime a thermometer so placed, and turned towards the clear sky, but away from the rays of the sun, will fall several de grees below the temperature of the surrounding air. 5. HYGROMETERS. Times of observation.-Observations of the dew-point hygrometer are as desirable at the regular hours as those of the other meteorological instruments; but, as more difficulty attends the observation, it is more liable to omission, and it is of great importance that when one experiment only can be made, the most advantageous hour should be selected for the purpose. Now it is probable that the minimum ·70 temperature of the alr in the 24 hours may correspond with the minimum temperature of the dew-point; and for the attainment of å mean result, the time of the highest dew-point should be selected, which would not differ much from 3 P.M., at which hour the observation should ou no account be omitted. The hygrometer should also be observed, if possible, at 9 A.M. and 9 P.M., but the minimum temperature might probably be substituted for the 3 A.M. observation without any material error. At the magnetical observatories the dew-point will be observed at the nearest magnetical hours to the meteorological hours above fixed on (3 and 9 A.M. and P.M.). Occasional observations of the dow-point under peculiar circumstances, as for instance in the inhabited apartments of houses or between the decks of the ships when laid up in their winter quarters in the polar regions, could not but afford information of high practical importance. All the ether of the dew-point hygrometer should be driven by the warmth of the hand from the covered ball into the uncovered, previously to an observation, and the ether should be dropped from a dropping-bottle very slowly upon the former. The temperature of the interior therinometer should be carefully noted upon the first appearance of the ring of dew upon the black bulb, and also its temperature upon its disappearance: the mean of the two obscrvations, should they differ, may be entered as the dew-point, together with the temperature of the air by the exterior thermometer. The wet-bulb hygrometer can be observed without difficulty, by mere inspection, and the observation should never be neglected at the regular hours. It is probable that the temperature of evaporation thus ascertained may afford the means of accurately determining the dew-point, and of solving all the points of hygrometry; but until all the necessary corrections shall have bevis agreed upon, one of the most essential requisites must be its frequcnt and accurate comparison with the dew-point, directly ascertained. The instruments supplied to the observatories are fitted with waterholders for keeping the bulb of the thermometer moist; but it should be observed, that the floss-silk, after being some time in use, has its faculty of conducting moisture greatly diminished, if not destroyed, and therefore requires to be renewed from time to time. When proper time can be afforded to allow the thermometer to take up its stationary temperature, it is, perhaps, best to moisten the ball for each obscrvation. By urging upon the wet bulb a jet of air from a small double-bellows, the maximum depression may be obd tained in a few seconds. In fact, the formula f"=f'- р 88 30 investigated by Dr. Apjohn, was tested under these circumstances, and he is of opinion that there will always be a small difference between the depression occurring in air at rest, and sweeping by the thermometer in a rapid current. In this formula, f" is the tension of steam at the dew-point; f' its tension at the temperature of the hygrometer; d the depres- . Х ·71 sion, or difference between the hygrometer and the air ; 88 a coefficient depending upon the specific heat of air and the caloric of elasticity of its included vapour; p the existing and 30 the mean pressure of the air. The hygrometers should be placed in the observatory, ncar to the standard thermometer, with which they should be frequently compared. 6. VANES, ANEMOMETERS, AND RAIN GAUGES. The magnetic observatories and the Antarctic Expedition have been furnished with Osler's self-registering anemometer and raingauge. In this instrument the direction of the wind is obtained by means of the vane attached to the rod, or rather tube, that carries it, and consequently causes the latter to move with itself. At the lower extremity of this tube is a small pinion working in a rack, which slides backwards and forwards, as the wind moves the vane; and to this rack a pencil is attached, which marks the direction of the wind on a paper ruled with the cardinal points, and so adjustod as to progress at the rate of 1 inch per hour, by means of a clock; the force is at the same time ascertained by a plate 1 foot square, placed at right angles to the vane, supported by two light bars running on frictionrollers, and conımunicating with three spiral springs in such a way that the plate cannot be affected by the wind's pressure without instantly acting on the springs, and connunicating the quantum of its action by a wire passing down the centre of the tube, to another pencil below, which thus registrrs its degree of force. The rain is registered at the same time by its weight acting on a balance, which moves in proportion to the quantity falling, and has also a pencil attached to it, recording the results. The receiver is so arranged as to discharge every half inch that falls, when the pencil again starts at zero. It is probable that the results obtained with this instrument would require correction for the varying effects of the eddy, which must be formed behind the board before they can be considered as exact measures of the pressure; and the effects of variations of temperature upon the force of the springs should be experimentally ascertained, particularly in very cold climates. This latter point may be determined by measuring the compression directly by the application of known weights. Another self-registering anemometer has recently been constructed by Professor Whewell, which exhibits upon a diagrain not only the direction and force, but the direction and integral effect of the wind, but which is more complex in its construction, and practically more liable to derangement. In it a small set of windmill vanes, something like the ventilators of windows, are presented to the wind by a common vane, in whatever direction it may blow. The current, as it passes, sets these vanes in rapid motion, and a train of wheels and pinions reduces the ·72 motion, which is thente communicated to a pencil traversing vertically, and pressing against an upright cylinder, which forms the support of the instrument: 1000 revolutions of the fly only cause the pencil to descend oth of an inch. The surface of the cylinder is covered with white paper, and the pencil, as the wane wavers, keeps tracing a thick irregular line, like the shadings on the coast of a map. The middle of the line may be easily traced, and it gives the mean direction of the wind, while the length of the line is proportional to the velocity of the wind and the length of time during which it blow's in each direction. Those who do not possess a register-anemonieter may make use of the common vanc and Lind's wind-gauge. The position of the former should be clear of all deflections and eddies from objects of the same or a higher level, and of course its position with regard to the true north should be clearly determined. In registering the direction of the wind it may be sufficient to use only 16 points of the compass. Lind's wind-gauge for measuring the force or momentum of the wind is adjusted for use by filling it with water till the liquid in both legs of the siphon corresponds with the 0° of the scale. It is to be held perpendicularly, with the mouth of the kneed tube turned towards the wind, and the amount of the depression in one leg, and that of its elevation in the other, are to be carefully noted. The sum of the two is the height of a column of water which the wind is capable of sustaining at the time, and every body that is opposed to that wind will be pressed upon by a force cquivalent to the weight of a calumn of water, having its base equal to the surface that is opposed, and its height equal to the altitude of the column of water sustained hy the wind in the wind-gauge. The height of this column being given, the force of the wind on square is casily found by a table which will be given in the Appendix. The observation of the gauge should always be made at the same point of a free space, and in gusty weather the maximum of the oscillation recorded. The most proper periods will be those of the other regular observations; but in great storms, or under other particular circumstances, occasional observations should be made, and registered apart. Even in observatories which are provided with Osler's apparatus it is desirable that an accurate comparison should be made of the two anemometers. The points most important to remark respecting the wind are, Ist. Its average intensity and general direction during the several portions of the day devoted to observation. 2ndly. The hours of the day or night when it commences to blow from a calm, or subsides into one from a breeze. 3rdly. The hours at which any remarkable changes of its direction take place. 4thly. The course which it takes in veering, and the quarter in which it ultimately settles. a foot ·73 5thly. The usual course of periodical winds, or such as remarkably prevail during certain seasons, with the law of their diurnal progress, both as to direction and intensity; at what hours, and by what degrees they commence, attain their maximum, and subside; and through what points of the compass they run in so doing. 6thly. The existence of crossing currents at different heights in the atmosphere, as indicated by the courses of the clouds in different strata. 7thly. The times of setting-in of remarkably hot or cold winds, the quarters from which they come, and their courses, as connected with the progressive changes in their temperature. Sthly. The connexion of rainy, cloudy or fair weather, with the quarter from which the wind blows, or has blown for some time previously. The Rain-gauge may be of very simple construction. A cubical hox of strong tin or zinc, exactly 10 inches by the side, open above, receives at an inch below its edge a funnel, sloping to a small hole in the centre. On one of the lateral edges of the box, close to the top of the cavity, is soldered a short pipe, in which a cork is fitted. The whole should be well painted. The water which enters this gauge is poured through the short tube iro a cylindrical glass vessel, graduated to cubic inches and fifths of cubic inches. Hence one inch depth of rain in the gauge will be measured by 100 inches of the graduated vessel, and Toboth inch of rain may be very easily read off. It is very much to be desired that, being of such easy construction, more than one of these gauges should be erected, or at least one placed with its edge nearly level with the ground, and all other upon the top of the highest building, rock, or tree in the immediate vicinity of the place of observation, the height of which must be carefully determined; it having been satisfactorily ascertained that the height of the gauge above the ground is a very material element in the quantity of rain which enters it. The quantity of water should be daily measured and registered at 9 A.M. 7. CLOUDS AND METEORS. Many very highly-interesting observations may be made, without the aid of instruments, upon the clouds., In describing them Mr. Howard's nomenclature may be adopted with great advantage. By means of the clouds different simultaneous currents of wind may often be detected, the different directions of which should be carefully ascertained by referring their motions to some fixed object. Their gradual evaporation or precipitation should also be carefully noted, and particularly their regular disappearance at night, or their more irregular and sudden formation. Rainbows, parhelia, haloes, &c., will of course be noted amongst the occasional remarks of the register; and an attempt should be made to express approximatively by numbers, the proportion which the overcast portion of the sky may bear to the clear space. For ·74 this the hemisphere Snay be supposed to be divided into eight sections, and the cloudy portion may be expressed by the fraction fth or kths, &c. 8. ELECTROMETERS. The Committee are fully impressed with the high importance of regular observations on the electrical state of the atmosphere; but they are not prepared to suggest any means of effecting this desirable object, which will at all correspond with the present advanced state of electrical physics. At no distant period they hope to see supplied a defect which is certainly a reproach to science. In the meantime much valuable information might be acquired by observations of an electroscope, on one of the ordinary constructions connected with a lofty insulated wire. In erecting such a wire, proper precautions should be taken against accidents by preparing a sufficient conductor in its immediate vicinity, by which a communication could be at once opened with the ground in case of any sudden and dangerous accumulation of the electric Huid. As a temporary contrivance, a common jointed fishing-rod, having a glass stick well varnisheil with shell lac, substituted for its smallest joint, may be projected into the atmosphere. To the end of the glass must be fixed a metallic wire terminating in a point, and connected with an electroscope by means of a fine copper wire. If the wire be made to terminate in a spiral'wrapped round a picce of cotton dipped in spirits of wiue and inflamed, its power of collecting electricity, will be sometimes doubled, but great precautions are necessary When this mode is employed. When the electroscope has been 'charged, the nature of the electricity may be tested in the usual way by excited glass or stuling wat The principal electroscopes which are capable of being employed to ascertain the electrical state of the atmosphere, or rather to compare its state at any given elevation with the state of the mediuin in contact with the instrument, are the following. 1. De Saussure's Electroineter, which consists of two fine wires, each terminated by a small pith ball, and adapted to a small inetal rod fixed in the upper part of a square glass cover, upon one of the faces of which a divided scale is marked, in order to measure the angles of deviation of the two balls. 2. Volta's Electrometer, formed of two straws about 2 inches long and {th of a line broad, suspended froin two small very moveable rings adapted to a metal rod : to measure the deviation of the straws a telescope with a nonius is employed. 3. Singer's Electrometer, consisting of two slips of gold leaf suspended from the rod. 4. Bohnenberger's Electroscope, formed of a single strip of gold leaf suspended from the conducting rod between two dry piles, the negative pole of one and the positive pole of the other being uppermost: this arrangement has the advantage of indicating the kind of electricity communicated to the conductor.

·75 The observations made with these and similar instruments have demonstrated that in serene weather the electricity of the atmosphere is always positive with regard to that of the carth, and that it becomes more and more positive in proportion to its elevation above the earth's surface; so that if an observer be on a mountain or in a balluon, if his conductor be directed downwards to reach an inferior stratum of air, his electroscope will indicate negative electricity; and if it be sent upwards into a superior stratum, positive electricity will be manifested. Various means have been resorted to in these experiments, such as connecting one of the extremities of the conducting wire to a kite, a small balloon, or the head of an arrow, the other extremity remaining attached to the electroscope. It has been ascertained by the observations of De Saussure, Schubler, Arago and others, that the positive clectricity of the atmosphere is subject to diurnal variations of intensity, thcre being two maxima and two minina during the twenty-four hours. The first minimum takes place a little before the rising of the sun ;• as it riscs, the intensity, at first gradually and then rapidly, increases, and arrives at its first maximum a few hours after. This excess diminishes at first rapidly and afterwards slowly, and arrives at its minimum some hours before sunset; it re-ascends when the sun approaches the horizon, and attaius its second maximum a few hours after, then diminishes till sunrise, and proceeds in the order already indicated. The intensity of the free electricity of the atmosphere has also been found to undergo annual changes, increasing from the month of July to the month of November inclusive, so that the greatest intensity winter, and the least in suinmer. In cloudy weather the free electrieity of the atmosphere still positive. During stornis, or when it raius or snows, the electricity is sometimes positive and sometines negative, and its intensity is always niuch more considerable than in serene weather. The electroscope will, during the continuance of a storm, frequently indicate several changes, from positive to negative. The above is a short summary ofalmost all that is known respecting the laws of atmospheric electricity. It will be highly important to obtain a series of observations equal in accuracy to those made by Schubler at Frankfort in 1811 and 1812, simultaneously with the observations of the hygrometer, barometer, thermometer, &c. Combined observations at a number of different stations canncs fail to give us important information respecting the distribution of the free electricity in the atmosphere, and the extent and nature of the disturbances to which it is subject; but to render the results valuable it will be necessary to have instruments comparable with each other, and this may be a difficult matter to effect *. Very recently a new method of investigating the electric state of the atmosphere has been proposed, likely to lead hereafter to very certain and valuable results; but it has not been sufficiently put in * For a fuller account of what is known respecting atinospheric electricity, and the mode of conducting the observations, see Becquerel's Traité de l'Electricité, t. iv. pp. 78_125. occurs ·76 practice to enable the Committee to recommend, at the present moment, the best form of instrument for making simultaneous and comparable observations, or the proper precautions to guide the observer in manipulating it. For the principle of this instrument we are indebted to Mr. Colladon of Geneva. He found, that if the two ends of the wire of a galvanic multiplier, consisting of very numerous coils well insulated from each other, were brought in contact, one with a body positively, and the other with a body negatively charged, a current of electricity passes through the wire, until equilibrium is restored; the energy and direction of this current is indicated by the deviation of the necdle from the zero-point of the scale. This instrument is applied to the purpose of ascertaining and measuring tne atmospheric electricity, by communicating one end of the wire with the carth, and allowing the other to extend into the region of the atmosphere, the clectrical state of which is intended to be compared. Thunder storms of course, should be attended to; but it is of consequence also to notice distant lightning not accompanied with thunder audible at the place of observation, especially if it take place many days in asuccession, and to note the quarter of the horizon where it appears, and th extent wich it embraces. In an actual thunder storm, especial notice should be taken of the quantity of rain which falls, and of the fits or intermittences of its fall, as corresponding, or not, to great bursts of lightning, as also of the direction of the wind, and the apparent progress of the storm with or against it*. 9. REGISTERS. The Register proposed by the Committee is comprised in two *skeleton forms, which have been supplied to the magnetical observatories and to the Antarctic Expedition. This, it will be observed, applies to the met corological forms about to be specified, independont of the inore extensive and general forms for the registry of the maguctic observations, with their accompanying meteorological entries, with which the magnetical observatories will be supplied. They are each calculated for one month's observation. The first form is for the insertion of observations as they are made in their uncorrected state. It consists of 12 p...cipal divisions, and is ruled across for 31 days, and for the arithmetical convenience of casting up the sums and means of the quantities inserted. At the bottom of the sheet there is also a space provided for the hourly observations of the baromcter and thermometers on the twenty-first day of the month, which wili be more particularly described after the explanation of the principal divisions. The outside compartments, both on the left and right of the sheet, are for the date of the month and the phases of the nioon. The second compartment is for the height of the barometer, and the temperature of the mercury for the four regular periods of observation, * On these subjects the Committee especially recomwend the attentive perusal of Arago's Notice sur le Tonnerre.

·77 The third compartment is appropriated to the dew-point hygrometer, and contains also four columns for the four daily observations, each of which is subdivided into three; for the temperature of the air, the dew-point, and the difference between the two. The fo rth compartment is for the wet-bulb hygrometer, and is similarly divided and subdivided for the temperature of the dry-and wet-bulb thermometer, and for their differences. The fifth compartiment is prepared for the maxima and minima of temperature, and is divided into three. In the first division are to be recorded the maxima and minima of thermometers carefully placed in the shade and screened from radiation. In the second, the maxima of a blackened thermometer exposed to the sun, and the minima of a thermometer placed in a inetallic mirror, and radiating freely to the clear sky. The third is devoted to occasional observations of the antinometer under favourable circumstances. The sixth compartment is for the temperature of the surface-water of the sea, or of any river in the immediate neighbourhood of the observatory The seventh compartment is prepared for observations upon the direction and force of th winar une four regular hours of registry. In the left-hand column of each dission is to be recorded the direction of the vane, and in the right- and colunu the height of Lind's gauge, in tenths of an inch of water. In the eightli compartment the amount of rain is to be registered once in the day; and su the ninth, the clectrical state of the atmosphere, if possible, at the four periods, 3 A.M., 9 A.M., 3 P.M., and 9 P.A. The tenth compartment is appropriated to remarks on the Houds, and weather generally; and in the eventh is to be noted, at hoone the longit le and latitude at sca. On a careful review of the month's observations, the maxima and minima results should have the algebraic signs + and respectively affixed. The second form is devoted to the corrected result of the observations, and to the optical comparison together of some of then, by their projection upon a scale of equal parts. The upper half of the sheet. verti. Il divideu i' 2 vo equal parts, each prepared for half the month's observations, and accordingly ruled across into sixteen spaces for the daily observations, and two for the suns and means of the quantities. Each half is also divided into five compartments. The first is for the date of the inonth and the phases of the moon. The second for the corrected height of the barometer at 32° Fahr. The third is appropriated to the elastic force of the aqueous vapour corresponding to the dew-poiut, and which may be taken from Table V. in the Appendix, p. 89. The fourth is for the maxiinum and minimum of temperature, and the mean of the two. And the fifth for occasional remarks. The lower half of the sheet is also vertically divided into two equal - ·78 parts, each of which is similarly divided into 31 columns for the daily observations of a month; and these again subdivided into four, for the six-hourly obscrvations of each day. The vertical lines thus formed arc divided into 6 inches; and each inch into tenths of an inch, and half-tenths, by horizontal lines. The left-hand compartment thus ruled, is intended for the projection of curves of temperature; for this purpose cach tenth of an inch upon the scale must be reckoned a degrce, which will be divided by the faint line into halves. The value of the degree may be arbitrarily fixed, and inserted in the margin according to convenience. Towards the upper part of the scale the results of the six-bourly observations should cach be marked by a dot in its appropriate space, and the dots may be afterwards connected by a line. The temperatures of the dew-point, or of the wet-bulb thermometer, or the mean temperature, may be compared with this primary result by projecting their curves in a similar way beneath it; and should the observations of these points be less frequent than four tiines in the day, the daily spaces inay casily be divided accordingly. The right-hand compartment is appropriated to the projection of cirves of pressure, and the four daily observations of the barometer are to be marked by dots towards the upper part of the scale of inches, and afterwards connected by a linc. Towards the lower part of the scale the elastic force of the vapour is to be noted, and the marks to be similarly connected by a line. On either the scale of temperature or of pressure, occasional comparisons may he made with results obtained at other stations, whichfif judiciously selected, cannot fail to prove of high interest and inuportance. They should, however, be laid down in pencil, or marked by a fainter lin. At the bottom of the first skeleton form will be found a space prepared for the 24 hourly obscrvations of the troenty-first day of the month, both in their uncorrected and their corrected state. It is divided into four compartments for 6 hours each. The instruments which can with most facility be observed in this manner, are the barometer with its attached thermometer, and the dry- and wet-bulb thermometers; and columns are appropriated to each of these. It is desirable that the means of cach 6 hours should be calculated, and spaces have been provided accordingly for the arithmetical operations. In casting up the sums and calculating the means, care should be taken in all cases tu verify the results by repetition; and the Committee recommend in every instance, before adding up the columns, to look down cach to sce that no obvious error of entry (as of an inch in the barometer, a very common error) may remain to vitiate the mean result. The precaution should also be taken of counting the days in each column, so as to make no mistake in the divisor. The skeleton forms will be interleaved with blank pages, to facilitate computations and comparisons, and to afford space for other observations of atmospheric phenomena, which will perpetually pre- ·79 sent themselves to those who make it their business or their pleasure to watch the changes of the weather on a judicious plan. The Committee, indeed, wish it to be understood, that, in the suggestions which they have offered, they have taken into consideration only such observations as are indispensable for laying the first foundations of meteorological science; some investigations of a more refined character they may, probably, make the subject of a future report. As soon as the register of a month's observations has been computed, it should be copied, and the copy carefully compared with the original by two persons, one reading aloud from the original, and the other attending to the copy, and then exchanging parts,-a process always advisable whenever great masses of figures are required to be correctly copied. A copy so verified should be transmitted regularly to such person or public body, as, under the circumstances, may be authorized or best adapted to receive and discuss the observations.

·No text on this page.

·81 APPENDIX. TABLES REFERRED TO IN THE SECTION ENTITLED “ INSTRUCTIONS FOR MAKING METEOROLOGICAL OBSERVATIONS." TABLE I. Correction to be added to Barometers for Capillary Action. Correction for Diameter of Tube. Unboiled Tubes. Boiled Tubes. inch. 0.60 0.50 0.45 0.40 0.35 0.30 0.25 0.20 0.15 0.10 inch. 0.004 0.007 0.010 0.014 0.020 0.028 0.040 0.060 0.088 0.142 inch. 0.002 0.003 0.005 0.007 0.010 0.014 0.020 0.029 0·044 0.070 ·82 TABLE II. Correction to be applied to Barometers with brass scales, extending from the cistern to the top of the mercurial column, to reduce the observation to 32° Fahrenheit. Inchcs. Temp. 20 20-5 21 21.5 22 22.5 23 23.5 + + 8 051 049 (148 046 044 042 -040 039 037 -035 033 + .053 .051 ·049 047 -045 -043 :042 040 •038 -036 034 054 -052 050 048 046 044 042 041 •039 •037 035 5 6 7 8 055 053 -051 .049 047 ·045 044 042 ·040 •038 -036 + 056 054 052 .050 048 046 •044 -042 041 039 •037 -058 -0.56 ·054 052 050) 048 046 + 059 057 055 053 .051 -049 047 044 .042 010 •038 + 060 058 056 -054 052 .050 -048 -046 013 •011 ·039 041 ·039 .037 10 0:14 •035 . •031 ·030 -028 026 024 022 021 ·019 •017 -032 ·030 -029 027 -025 11 12 13 14 15 16 17 18 19 20 035 •033 031 029 -027 *033 ·031 -029 •027 -026 .024 -022 020 •018 016 030) 028 ·026 .024 022 -020 018 -016 .036 ·034 •032 030 -028 -026 024 -022 -020 •018 •031 -029 -027 025 -023 021 ·019 •017 •037 .035 -033 031 ·029 ·026 024 021 ·019 ·018 -016 023 021 ·019 -017 020 ·018 21 22 014 012 010 •(N18 006 ·005 014 -012 010 008 ·007 ·005 003 ·001 011 012 010) -009 -007 ·005 -003 ·001 015 -013 •011 -009 007 005 -003 00)1 015 013 •011 .009 -007 0801.3 -003 ·001 -015 ·013 •011 009 ·007 0305 .003 015 ·013 •011 009 007 ·005 003 016 014 •012 ·010 ·007 25 26 27 28 003 •0303 .00) 001 -001 -001 29 30 001 003 001 003 001 -003 ·001 003 001 003 001 ·003 -001 ·003 -001 -003 31 32 33 34 35 36 37 38 39 40 ·005 .006 ·008 •010 012 •013 ·015 ·017 019 -021 ·005 ·006 008 -010 012' ·014 •016 017 019 021 ·005 .007 008 010 012 ·014 •016 ·018 ·020 022 ·005 -007 009 •011 013 014 ·005 ·007 •009 011 -013 015 ·017 019 021 -023 ·005 •007 009 •011 -013 015 ·017 019 021 023 ·005 007 009 011 013 •016 ·018 020 022 -024 ·005 -007 ·010 •012 -014 •016 ·018 -020 -022 024 *016 018 ·020 41 42 43 026 45 46 47 48 49 50 •U23 -025 027 ·029 •030 032 034 036 038 039 ·030 031 *033 035 037 038 024 025 027 -029 031 033 035 037 039 040 024 -026 .(28 ·030 •032 -034 036 038 040 041 -025 -027 ·029 031 •(133 -035 036 038 040 042 ·026 -028 -030 032 *034 036 038 -025 •027 ·029 •031 033 •035 037 ·039 041 -043 ·026 028 •031 •033 035 •037 •(39 041 043 045 040 042 044 ·83 TABLE II. (continued). Inches. Temp. 20 20-5 21 21.5 22 22.5 23 235 010 042 .046 0:12 61 52 53 54 55 56 57 58 59 60 -0,16 017 019 ·031 -033 -0.55 -0.56 041 043 045 017 -019 -0.50 032 -0,34 (036 •058 •016 0448 .050 .052 054 053 ·057 -059 -013 015 •017 019 -0.51 053 .055 .057 ·059 -061 016 •018 .050 052 •0)51 056 -058 .060 ·002 045 -047 049 .051 053 055 •037 ·059 •061 063 018 .050 •052 .055 057 059 001 063 065 047 049 052 054 .056 058 060 062 064 066 063 #067 61 .(38 62 .060) 63 •002 61 ·063 65 ·065 60 067 67069 66 .071 69 -072 70 •071 060 ·001 063 .063 .067 .069 •071 •07: 074 •076 .061 063 •1065 067 068 •070 •072 -07- 076 078 -062 061 .066 .068 070 •0722 071 076 -078 080) 061 .066 068 -070 •072 07. ·009 •071 073. •075 •077 ·079 .081 •083 -067 009 *071 073 075 077 ·079 ·081 083 085 068 070 •072 075 -077 ·079 •081 083 085 •087 •078 •080 71 72 73 74 75 76 77 78 79 80 076 •078 (179 081 •083 •083 •087 088 090 •078 •080 •08) •083 -083 •087 •089 -091 •080 .082 -083 •083 •087 •089 091 09:3 082 031 085 •087 089 091 .(03 083 085 087 089 091 •093 09.5 .(197 -0.99 101 085 087 089 ·091 093 ·095 *097 099 •101 103 •087 089 091 093 -093 •097 .100 .102 •101 089 091 093 ·095 098 •100 •102 •104 .106 •108 •096 097 ·099 •106 094 -0.93 098 . 100 •097 81 82 83 84 85 86 87 88 89 90 .099 •101 •103 .104 106 •108 •110 096 038 1080) •101 .103 •103 •107 •109 •111 •112 .101 •106 •108 •109 •111 .113 .115 •101 103 •104 106 . 108 110 .112 •114 116 •118 •103 •105 •107 •109 111 •113 .115 •117 •119 .121 •1005 -107 •109 111 •113 •115 •117 •119 •121 •108 •110 112 114 •116 •118 .120 •122 •1244 .126 .110 •112 •114 •116 118 120 •123 .125 •127 .129 91 92 93 94 95 96 97 98 99 100 •111 113 •115 •117 118 .120 122 •124 .125 •127 111 116 •118 •120 •121 •123 123 •127 .129 .130 •117 •119 .121 122 •1:24 •126 •128 •130 •132 •134 120 122 •124 •125 •127 .129 •131 •133 •135 .137 -122 .124 •126 •128 •130 .132 .134 •136 .138 •140 125 •127 129 •131 •133 .135 •137 •139 •141 143 •128 •130 •1:2 •131 •136 •138 .140 • 142 114 146 .131 •133 •135 137 .139 •141 •143 115 •147 •150 G 2 ·84 TABLE II. (continued). Inches. Temp. Temp. 24 21.5 25 25.5 26 26-5 27 27.5 + + 069 .067 2 3 + .061 059 057 055 .053 051 049 046 044 042 040 + ·063 061 .058 •036 .(054 -052 050 .047 045 ·043 041 + ·064 .062 .060 .057 055 053 051 048 ·046 044 .042 5 + -065 •063 -061 ·059 -056 054 052 •049 047 ·045 .042 + 067 ·064 062 060 057 055 -053 050 .048 046 -043 + .068 065 •063 •OG1 -058 056 .054 -051 049 .046 044 3 4 5 •071 068 ·066 063 061 -058 -056 053 051 048 046 002 ·059 -057 055 052 050 047 015 7 7 8 9 10 9 10 11 12 13 •042 ·039 037 -035 11 12 13 15 16 17 18 19 20 038 036 •033 •031 029 ·027 023 -023 021 •018 ·039 •036 034 •032 •030 028 025 -023 -021 019 •039 037 •035 •033 030 •028 (26 024 -021 ·019 ·040 •038 036 033 •031 ·029 020 -024 022 -020 .041 •039 -(136 034 •032 ·029 027 ·025 ·022 -020 042 040 -038 035 •033 ·030 -028 (925 -023 021 -043 041 038 036 •033 031 ·028 -026 024 021 ·030 .027 -025 -023 020 15 16 17 18 19 20 016 •017 •014 21 22 23 24 25 26 27 28 ·014 010 ·008 ·005 -003 001 014 ·012 010 .008 006 •003 ·001 -017 •015 ·012 •010 -008 ·006 ·003 001 017 015 •013 .010 •008 006 ·003 001 -018 -015 ·013 •011 -008 •006 004 001 ·018 •016 •013 •011 ·008 006 -004 ·018 •016 ·013 •011 ·009 006 ·004 ·001 -019 016 •014 011 -009 .006 -004 -001 21 22 23 24 25 26 27 28 29 30 ·001 003 001 003 ·001 .003 001 .004 001 ·004 001 004 ·001 ·004 001 004 29 30 DO 006 .006 31 31 32 33 34 35 36 37 38 39 40 -005 -008 010 •012 ·014 -016 -018 -020 023 -023 ·008 ·010 ·012 014 ·017 •019 021 .023 025 .006 008 •010 ·012 015 017 ·019 -021 024 026 -006 008 -010 •013 ·015 ·017 ·019 ·022 -024 026 ·006 .008 011 •013 ·015 ·017 ·020 -022 -024 ·027 ·006 008 •011 •013 ·015 .018 020 .023 -025 .027 006 008 •011 ·013 ·016 018 021 -023 -025 -028 009 011 •014 ·016 ·019 021 ·023 ·026 028 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 027 ·029 031 033 035 •038 040 042 044 046 -027 030 •032 034 036 038 041 043 •045 -047 028 030 032 •035 •037 039 -041 044 046 -048 029 031 •033 035 -038 ·010 .042 045 047 049 ·029 -031 -034 036 •038 041 ·043 045 048 050 •030 032 034 037 039 042 044 046 -049 051 030 033 035 •037 ·010 -042 -045 -047 050 052 031 033 •036 038 041 •043 046 048 050 053 41 42 43 44 45 46 47 48 49 50 ·85 TABLE II. (continued). Inches. Temp. Temp. 24 24.5 25 25.5 26 26.5 27 27-5 51 53 55 56 57 58 59 60 •018 050 053 -0.35 •057 059 061 -063 065 .068 049 052 054 (156 -058 OGO 062 065 067 069 •050 053 055 •057 059 061 ·064 066 .068 070 051 054 056 058 •060 ·063 065 067 •070 •072 052 .055 057 ·059 062 •064 .066 069 071 073 .053 056 058 -060 ·063 065 068 -070 072 075 054 •057 ·059 .062 064 066 069 .071 •074 •076 -055 058 060 -063 065 068 -070 073 -075 077 55 56 57 58 59 60 6) •070) •072 •074 076 078 -074 076 -079 •081 61 62 63 63 64 •071 •073 •076 •078 -080 •082 081 086 089 091 64 073 075 -077 •079 082 084 ·086 .088 090 093 075 078 080 082 085 .087 (189 -092 .094 096 •037 079 •082 •084 086 •089 091 094 096 098 •078 081 083 •086 .088 .090 -080 082 •085 •087 ·090 092 ·095 097 •100 •102 66 080 083 .093 65 66 67 68 09 70 085 088 090 092 093 68 69 •083 .087 (189) 095 098 100 70 71 72 73 74 75 76 77 78 79 80 091 093 095 097 •100 102 •104 •106 . 108 •110 •093 ·095 .097 099 102 •104 106 •108 •110 •113 095 097 ·099 •102 •104 •106 108 110 •113 •115 •097 ·099 101 •104 •106 •108 •110 •113 •115 •117 -093 .101 •103 106 •108 •110 •112 115 •117 •119 •301 •103 .105 •108 110 •112 •115 •117 •119 •122 102 105 •107 .110 112 •114 •117 •119 -122 •12+ 104 71 •107 72 -109 73 .112 •114 | 75 •117 76 •119 77 122 78 •124 79 •126 80 •126 81 82 83 •112 .114 •117 •119 .121 •123 .125 127 .129 •131 •115 •117 •119 .121 •123 • 126 •128 130 .132 •134 •117 .119 •121 124 •126 128 130 •133 •135 •137 .119 •122 •124 •126 •128 •131 133 •135 •137 •140 •122 .124 .126 •129 .131 •133 •136 •138 •140 142 •124 .126 •129 •131 .133 •136 •138 •141 •143 •129 •131 •134 •136 •139 •141 .143 •146 .148 •151 •129 •131 134 •136 •138 •141 •113 •146 •148 81 82 83 84 85 86 87 88 89 90 86 87 88 89 90 •145 91 92 93 94 95 96 97 98 99 100 •131 .136 •138 .140 •142 •144 •146 •148 •151 .153 136 139 •141 •143 145 •147 •149 •152 •154 •156 .139 •141 •144 •146 •148 •150 •152 .155 •157 •159 .142 144 •147 149 •151 •153 •156 .158 •160 •162 145 .147 •149 .152 .154 •156 •159 •161 •163 165 .148 •150 .152 .155 •157 •159 .162 •164 •166 •169 150 153 •155 •157 160 •162 •165 •167 169 •172 .153 •156 158 •161 .163 •165 168 •170 .173 •175 91 92 93 94 95 96 97 98 99 100 ·86 TABLE II. (continued.) Inches. Temp. Temp. 28 28.5 29 29.5 30 30-5 31 + + -- + 072 •069 •067 ·064 0 1 2 3 4 5 074 072 0 1 2 3 069 •073 071 .068 ·005 ·063 .060 -058 (155 •053 -050 •017 + 076 •073 070 •068 ·065 •062 060 •(137 ·054 052 049 067 .064 061 ·059 056 054 .051 048 077 -074 •072 .069 066 -063 061 -058 •055 -059 -037 •054 052 (149 •047 + •078 •076 073 -070 •067 ·065 062 059 •056 ·034 051 080 •077 -074 •071 -068 -066 ·003 .060 -057 054 052 7 8 9 10 5 6 7 8 9 10 CCO. OSO. 046 11 12 13 14 041 (012 039 •037 ·034 •032 ·029 ·026 -021 ·021 (145 0.12 040 •037 (35 .032 ·030 -027 ·02-1 .022 016 •043 ·040 038 •035 033 030 .027 15 16 17 18 19 20 •017 045 -042 -039 .036 •034 •011 •038 030 1733 031 ·028 ·025 023 048 0.15 013 ·010 -037 034 032 ·029 (126 .023 ·019 0.16 -0.13 -010 .038 -035 11 12 13 14 15 16 17 18 19 20 180. 870). ·026 023 -029 027 ·024 610- 913). 010). 410. 120. 810 21 22 23 91 25 26 27 28 •014 OL -009 •006 ·001 -(001 -020 ·017 ·014 •012 -009 2007 ·001 .(0)1 ·020 ·017 015 •012 •009 •007 ·001 ·001 -020) -018 (015 •012 -009 210. 600) 900). 100) 100) 021 018 ·015 ·012 ·010 015 013 010 007 .004 21 22 23 24 25 26 27 28 200 VOO. 100. 200. V00. 100. 100. - 29 30 001 -004 100). 100. ·001 -004 -001 ·001 .001 •004 .001 004 ·001 -004 29 30 .000 900) 600). 200. 600). 710. 10. 200. 600. 600 110 F10- 31 32 33 34 35 36 37 38 39 40 •017 ·020 ·012 •014 017 ·019 022 0244 -027 -029 -019 -021 007 009 -012 •015 •018 -020 023 ·026 .028 031 ·012 015 017 ·020 022 025 ·028 ·030 ·007 •010 •012 ·015 018 021 023 ·026 ·029 ·031 007 010 •012 ·015 •018 -021 -024 026 029 -032 31 32 33 34 35 30 37 38 39 40 024 026 -029 023 ·027 030 41 032 •034 0937 034 •036 039 -042 42 43 43 45 46 47 48 49 50 031 034 -036 -039 011 044 046 049 051 .054 -042 015 047 050 -052 055 033 035 038 010 043 •045 •048 051 053 056 033 036 •038 041 ·044 046 049 (152 054 -057 -047 050 052 055 058 034 -037 0.10 0:12 -045 •018 .151 053 056 059 .035 •037 040 ·013 -046 -049 .(051 054 057 060 45 46 47 48 49 50 ·87 TABLE II. (continued). Inches. Tenip. Temp. 28 28.5 29 29.5 30 30-5 31 .062 065 068 51 52 53 54 55 56 57 58 59 60 056 059 061 .064 -066 069 •071 071 076 ·079 -057 -060 ·063 065 068 -070 •073 075 078 -080 058 -061 064 066 069 •071 •074 •077 ·079 082 ·059 062 065 067 070 •073 .075 •078 080 083 060 063 -066 068 071 •074 •076 •079 061 064 -067 •070 072 075 -078 ·081 •083 086 51 52 53 54 55 56 57 58 59 60 085 61 62 63 64 65 66 67 68 69 70 081 084 086 089 ·091 094 .096 099 •101 •104 083 083 088 090 093 096 098 •10) •103 •106 .084 -087 089 092 095 097 •100 102 .103 .]08 •086 -088 •091 -0994 .096 ·099 •102 •104 •107 •109 -087 -090 093 095 098 •101 103 100 -109 •111 089 091 091 097 -100 102 •105 •108 •110 •113 090 -093 096 098 -101 •104 •107 •109 112 •115 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 71 72 73 74 75 76 77 78 79 80) •106 •109 •111 •114 116 •119 •121 •121 •126 129 108 111 •113 116 118 •121 .123 126 •128 .131 110 •113 .115 •118 .120 •123 .126 -128 .131 1:33 •112 115 •117 •120 .122 •125 •128 •130) .133 •130 •114 •117 •119 122 •125 127 •130 •133 •135 •138 •116 •119 •121 124 127 .129 132 .135 .137 • 110 •118 •120 •123 126 •128 •131 •134 •137 .140 •143 77 78 79 80 81 82 83 84 85 86 87 88 89 90 •131 .134 •136 •139 •141 •144 •146 .149 •151 •153 81 82 83 84 85 134 136 139 •141 •144 •1.16 149 •151 -154 •156 .136 .138 •141 •144 .146 •149 •151 .154 •156 .159 •138 141 •143 •146 •149 •151 154 •157 •159 •141 •143 •146 •149 •151 •15+ 157 .159 .162 •164 •143 •146 •148 151 151 •156 •159 .162 .165 ..167 •145 •148 •151 •154 •156 .159 86 87 88 .165 .167 •170 .162 89 90 91 92 93 94 95 96 97 98 99 100 156 158 •161 .163 .166 168 •171 •173 •176 •178 •159 •161 *164 •166 .169 •171 .174 •176 .179 .181 102 •164 167 •169 •172 •174 •177 •179 .182 .184 165 .167 •170 •172 .175 •178 .180 •183 •185 .188 •167 •170 •172 .175 •178 .181 .183 .186 .188 •191 •170 •172 •175 •177 •180 .183 .186 .188 •191 •194 •173 91 •175 92 .178 93 •180 94 •183 95 .186 96 •189 97 •191 98 •194 99 • 197 100 ·88 TABLE III. Correction to be applied to Barometers, the scales of which are engraven on glass, to reduce the observations to 32° Fahrenheit. Temp. Inches. Inches. Inches. Inches. Inches. Inches. Inches. Inches. 28.0 28.5 290 29.5 30.0 30.5 S10 31.5 25 30 35 40 45 50 +.017 +.017 +.017 +.018 +.018 +.018 +.019 +0.019 +.005 +.005 +.005 +.005 +.005 +.005 +.005 +.005 -007-007 - 007-008-008--008 -008 --008 - 019 -020 - 020 -020 --021 --021 -021 -.022 -031 --032 -032 --033 --033 -034 035 -.036 -043 --044 -1945 --046 --046 - 047 -.048 - .049 --055_056 - 057 --058 -.059 - 060-061-062 -.067 -068 -069 - 071 -072 - 074 -075 - 076 --079 -081 - 082 -083 -.085 -.086 -.088 -089 -091 -.093 -.094 -.096 -.098 -100 -•101 - 103 -]03 - 105 106 -•109 -.111 -114 116 – 118 55 60 65 70 75 - . TABLE IV. Showing the Force of the Wind on a square foot for different heights of the Column of Water in Lind's Wind-gauge. Height of the Column of Water. Force of the Wind in Avoirdupois Pounds. Common designation of such a Wind. inches. 12 11 10 9 8 7 6 5 } Most violeut hurricane. 62.5 57.29 52-08 46.87 44.66 36.55 31.75 26.04 20.83 11-62 10-42 5.21 2.60 0.52 0.26 A very great hurricane. A great hurricane. A hurricanc. A very great storm. A great storm. A storm. A very high wind. A high wind. A brisk gale. A fresh brecze. A pleasant wind. 1 0.5 0.1 0.05 In great degrees of cold, a saturated solution of sea salt may be used instcad of water, the specific gravity of which is 1.244. If the force in the above Table for any height be multiplied by the specific gravity, the product will be the true force, as measured by the solution.

·89 TABLE V. Elastic Force of Aqueous Vapour for every degree of Temperature, from 0° to 124° Fahr. Temp. Fahr. Force, Inches of Mercury. Temp. Fahr. Force. Inches of Mercury. Temp. Fabr. Force. Inches of Mercury. Temp. Fahr. Force. Inches of Mercury. od 1 2 3 4 5 6 7 8 9 10 32 33 34 35 36 37 38 39 40 41 42 43 11 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 ! 83 i 44 12 13 14 0.051 0.053 0.056 0.058 0.000 0.063 0.066 0.069 0.071 0.074 0.078 0.081 0.084 0.088 0.092 0·095 0.099 0.103 0.107 0.112 0.116 0.121 0.126 0.131 0.136 0.142 0.147 0.153 0.159 0.165 0.172 0.179 15 0.186 0.193 0.200 0.208 0.216 0.224 0-233 0.242 0.251 0-260 0.270 0.280 0.291 0-302 0.313 0.324 0.336 0.349 0.361 0.375 0.389 0.40) 0.417 0.432 0.447 0.463 0.480 0.497 00514 0.532 0.551 45 46 47 48 49 50 51 52 53 0.570 0.590 0.611 0-632 0.654 0.676 0-699 0.723 0.748 0.773 0.799 0.826 0.854 0.882 0.911 0.942 0.973 1.005 1.036 1.072 1.106 1•142 1.179 1.217 1.256 1.296 1:337 1.380 1.423 1.468 1.514 94 1.562 95 1.610 96 1.660 97 1.712 98 1•764 99 1.819 100 1.874 101 1.931 102 1.990 103 2.050 104 2.112 105 2.176 106 2.241 107 2-307 1 108 2.376 109 2.447 110 2.519 111 2.593 112 21169 113 2.747 114 2.826 115 2.908 116 2.992 117 3.078 118 3.166 119 3.257 120 3:349 121 3.444 122 3.542 123 30641 124 3.743 16 17 18 19 20 21 22 84 54 85 86 87 88 25 26 27 28 55 56 57 58 59 60 61 62 89 29 90 91 92 93 30 31

·No text on this page.

·91 The following letters from Baron Von Humboldt to the Earl of Minto, and from Professor Erman to Major Sabine, having been communicated to the Royal Society, the Committee have deemed it desirable that they should form a part of this publication, with the permission of the respective parties. Letter from the Baron Alexander von lluinboldt to the Earl of Minto. MILORD: Berlin, ce 12 Oct., 1839. LORSQ'AC printems de l'année 1836, j'aullessai une lettre à S.A.R. Mgr. le Duc de Sussex sur les moyens propres à perfectionner la connoissance du magnétisnie terrestre par l'établissement de stations inagnétiqucs et d'observations correspondantes, je sollicitai le concours puissant de la Société Royale de Londres on favcur de travaux qui, émanant à la fois de plusieurs grands centres scientifiques de l'Europe, pourroient conduire progressivement à la conndissance précise des lois de la nature. Ma démarche fut accueillie avec bienveillance, et la Société Royale daigna recomniander à la protectioli spéciale du Gouvernement de Sa Majesté l'établissement de plusieurs stations permanentes dans les régions tropicales, et dans les parties tempérées de l'hémisphère austral. Cette protection du Gouvernement a été accordée avec une munificence qui dépasse de bien loin l'espoir des lionunes le plus ardemment occupés des variations du magnétisme terrestre, selon les trois coordomécs de déclinaison, d'inclinaison, et d'intensité absolue. Ce ne sont pas seulement des stations magnétiques qui seront fondées dans les licux les plus propres à la inanifestation des changemens que subit la distribution des forces; c'est une grannie expédition untarctique qui a été ordonnée sous le commandement d'un savant et intrépide navigateur, le Capitainc James Clark Ross; expédition qui cnibrassera dans des travaux sagenent preparés tous les problèmes du magnétisme terrestre, de la configuration du globe, de la distribution de la chaleur, du mouvement des caux de l'océan, de la constitution géologique du sol, de la géographie des plantes et des aniinaux. Je crois remplir un devoir sacré en offrant au Premier Lord de l'Amirauté, à Monsieur le Comte de Minto, l'hommage respectucux de la plus vive reconnoissance dont sont pénétrés tous ceux qui cultivent les sciences, ct lcur ont voué une vie laborieuse. Cette re- ·92 connoissance est due au Ministre qui, dans des vues élevées et si favorables aux progrès de l'intelligence, a réalisé l'exécution du voyage antarctique. La bienveillance personnelle dont votre Excellence m'a honoré, pendant Son séjour à Paris et à la Cour de mon Souverain, me donne le courage de Lui communiquer en même tems quelques considérations qui se rattachent au but principal d'une vaste et noble entreprise. Ma franchise ne sera pas mal interprétée. La variabilité des phénomènes est ce qui caractérise le plus le inagnétisme terrestre : variabilité selon une marche lente et périodique, quelquefois intermittente aussi, comme effet de perturbations brusques et instantanées. Il en résulte que pour approfondir les lois du magnétisme terrestre, il est ci'unc haute importance de connoître l'état magnétique du Globe à une même époque donnée, ou, du moins, selon des obscrvatious faites à «les époques très rapprochées. Il y a presque déjà trente ans que, dans le Recueil de mes Observations astronomiques, j'ai indiqué combien il serait précieux pour la Physique du globic, si plusieurs bâtimens munis d'excellens instrumens, parcouraient simultanément l'équateur magnétique et les lignes sans déclinaison, pour fixer à la même époque, dans le vaste bassin des niers, la déclinaison, l'inclinaison, et l'intensité des forces magnétiques. J'insistai aussi, (malgré l'imperfection des instrumcns et des méthodes d'alors,) d'après ma propre expérience, sur la possibilité de déterminer sur mer, et avec une précision suffisante, les variations de ces deux derniers élémens. (Rel. hist. T. 1. p. 262.) Je montrai conubien ces déterminations océaniques senbloient offrir d'avantages, là, où les couches d'eau sont assez épaisses pour que l'on ait moins à craindre les perturbations locales dues à la constitution ipinérg.logique du foud. Guidé par des considérations analogues, j'ose exprimer le désir, que pour rendre plus fructueux encore l'immense travail qui sera exécuté en trois années, soit par l'expédition du Capitainc Ross, soit dans les nombreuses stations magnétiques répauducs sur la surface des continens et des iles, Votre Excellence voulůt bien ordonner simultanément quelques expéditions partielles et supplémentaires. Deux savans, auxquels nous devons des travaux inportaus sur la connoissance des variations du magnétisme terrestrc, M. le Major Sabine et M. Lloyd, professeur à Dublin, m'ont déjà donné l'heurcuse nouvelle que le Gouvernement de Sa Majesté enverroit à Otahiti, à cette métropole de l'Océan pacifique, illustrée par d'anciens travaux astronomiques, un officier très-instruit et muni d'appareils magnétiques. Le grand nombre de bâtimens de la marine royale qui se trouvent le plus souvent en station sur les côtes occidentales de l'Amérique du sud, et dans les mers de l'Inde, faciliteront peut-être les moyens de multiplier les investigations que j'appelle supplementaires, et dont, pour le moment, le hut principal seroit la connoissance expériinentale de l'équateur magnétique, ct des lignes sans déclinaison. I. Un bâtinent inuni d'instrumens propres à mesurer l'inclinaison, la déclinaison, et l'intensité, pourroit, en partant des côtes du Pérou, suivre l'équateur inagnétique, ou la courbe d'inclinaison zéro, jusqu' ·93 aux côtes de la péninsule de Malacca, et, si le vent le permet, jusq'au détroit de Bab-el-Mandeb. Un second bâtiment pourroit parcourir l'équateur magnétique depuis le Golfe de Guinée jusqu'aux côtes du Brésil. On détermineroit avec une grande précision astronomique les points du littoral où la courbe d'inclinaison zéro, qui n'est pas un grand cercle de la Sphère, coupe les continens et let iles: on apprendroit à connoître les changemens de sinuosité et le mouvement des nouds (points d'intersection des équateurs magnétique et terrestre) qui ont eu lieu depuis les époques des voyages antéricurs. Comnie les lignes isodynamiques et isoclines ne sont aucunement parallèles, il seroit à désirer que les intensités fussent aussi déterminées le long de l'équateur magnétique, ou dans sa proximité la plus immédiate. II. Quant aux parties des lignes sans déclinaison qui deviennent accessibles aux navigateurs, j'oserai, Monsicur le Comte, les indiquer toutes, nou dans le vain espoir que des observations simultanées puissent les embrasser dans leur ensemble pendant la durée du séjour du Capt. Ross dans les hautes régions antarctiques, mais seulement pour faciliter le choix à Votre Excellence selon les combinaisons fortuites que peuvent offrir des traversets ou les stations éphémères de bâtiniens de la marine royale. Je n'ignore pas que d'après les grandes vues sur les véritables fondemens d'une Théorie générale du magnétisme terrestre qui sont dues à M. Gauss, soit la connoissance approfondic de l'intensité horizontale, soit la multiplicité et la sage réparpartition des points dans lesquels les trois élémens de déclinaison d'inclinaison et d'intensité ont éte simultanément inesurés, pour trouver la valeur de V (§. 4. ct 27), et par conséquent aussi de sont les points vitaux du problênie qu'a résolu l'illustre Géomètre : mais les besoins actuels du Pilotage, les corrections habituelles du runıb, et des chemins parcourus, donnent encore une importance speciale: ct pratique à l'élément de la déclinaison. On apprécirroit une détermination expérimentale, c. d. d. par observation inimédiate, avant que l'édifice théorique ait pu être complété et terminé dans son ensemble; on l'apprécieroit d'autant plus que les lignes isogones ont un mouvement très-inégal dans les différentes portions de leurs traces, et que l'action combinée des petites attractions magnétiques locales cause des déviations partielles de la direction inoyenne des lignes d'égale déclinaison, déviations qui intéressent la sécurité des routes, et qui resteront long-temps hors de l'atteinte de la théorie générale la plus solidement établie. Je signale ici de préférence la direction des lignes sans déclinaison, auxquelles des considérations de Géographie physique doivent conserver une partie de leur ancienne importance. (a.) L'expédition antarctique, en arrivant, par l'ouest, de la Terre Kerguelen à celle de Van Dieinen, aura traversé la ligne sans déclinaison qui remonte au nord vers la Terre de Nuyts (Australie.) Il seroit important de fixer astronomiquement, comme je l'ai fait observer pour l'équateur magnétique, les points méridionaux et septentrionaux du littoral de la Nouvelle Hollande, où la ligne de déclinaison zéro ·94 traverse le continent australien, et de poursuivre cette courbe, d'abord vers l'O. N. (), ct ensuite vers le nord, depuis la Baie de Vansittart, ou le Cap Bougainville, jusq' aux iles Maldives, et les attérrages de Surate dans l'Inde. Les connoissances acquises par les beaux travaux de Hansteen, d'Adolphe Ernan et de George Fuss sur la grande sinuosité des ligues isogones de la Sibérie enipéchent aujourd'hui de se former une idéccexacte de la liaison de ces lignes avec les lignes correspondantes dans les Mors de l'Inde et de la Chine. D'après les cartes intéressantes qui accompaguent l'exposé de la Théorie générale par M. Gauss, la ligne de déclinaison zéro ne coupe le continent asiatique que près de l'entrée du Golfe Persique ; cile remonte directement de là vers le nord à la Mer Caspionne et à la Mer Blanche. D'après M. Barlow clle se replie du Golfe de Cambaye vers le N.E. ct reparoît dans les Mers de la Chine et du Japon, entre l'extrémité septentrionale de l'Ile Forinose et la péninsule Seghalienne. Ce seroit jeter une vive lumière sur un des points les plus obscurs du magnétisme terrestre quc de lever les doutes qui cnveloppent le prolongement de cette ligne de cléclinaison zéro de la iner des Indes, et de faire connoître, par des observations précises, la direction et la distribution des forcés à l'ouest de l'Indus entre Candahar, Balkh, Koundouz, et le Pendjab (la Pentapotamie). Il est probable que la marche victoricuse des arınées de S. M. vers Caboul, et le séjour des troupes dans l'Afghanistan pourront donner lieu à des recherches de ce genre, au moyen des petits appareils magnétiques que l'on destine pour l'Inde. Il resteroit à examiner pour la même époque, la position de la ligne zéro dans les mers du Japon au nord de l'Isla Formose, comme dans l'Océan Glacial dans la partie trèsaccessible entre Spitzberg et la Mer Blanche. Suivre les traces de l'équateur magnétique, ou celles des lignes sans déclinaison, c'est gouverner (diriger la route du vaisseau) de manière à couper les lignes zéro dans les intervalles les plus petits, en changeant de rumb chaque fois que les observations d'inclinaison ou de déclinaison prouvent qu'on a dévié. (1.) Si du système oriental, ou de l'ancien continent, nous passons au système magnétiquc americain et atlantique, nous aurions à désirer la détermination simultanée des portions de la ligne sans déclinaison qui remonte à l'est de la Géorgie du sud vers St. Salvador am Brésil, quitte le continent près de Maranham, ct se dirige au N. O. vers le Cap Charles et la Baye de Chesapeak. Les mers que traverse cette ligne sont si fréquentécs que de nonıbreuses observations magnétiques y ont été faites, et se trouvent conservées dans les archives du Dépôt de la Mariue Royale; mais il ne suffit pas d'avoir coupé souvent et à différentes époques la ligne zéro, il s'agit de la poursuivre, autant que les vents le perniettent, dans toute son étendue. Je devrais hésiter, M. le Comte, à faire mention du prolongement le plus boréal de la ligne atlantique à travers le Canada et la Baie d'Hudson, mais je dois considérer la surface du globe dans son ensemble, et fixer l'attention des navigateurs sur les changemens qui peuvent être survenus dans les dernières anuées.

·95 (c.) La Mer du Sud, si l'on en excepte les côtes du Japon n'a de nos jours pas de variation zéro. Le noud circulaire qui renferme l'archipel des Marquesas près du minimum des variations orientales (59) mérite de nouvelles investigations dont pourroit se charger le bâtiment qui suivrait l'équateur magnétique du Pérou vers l'Inde. La forme de ce neud circulaire c. à d. l'espacement, variable des courbes isogones qui le constituent, et le déplacement progressif du naud cntier, sont des phénomènes également remarquables et qui contrasteut avec le grand noud circulaire de l'Asie orientale, auquel, selon le mémoire de M. Gauss, appartient la courbe de déclinaison zéro des mers du Japon et de la Chine. Je compte sur votre ancienne bienveillance, Milord, en osant vous importuner si longuement de considérations sur l'utilité que pourraient offrir des observations simultanées, par l'emploi d'instrumens et de méthodes semblables, dans les différentes régions des deux hémisphères. J'ai touché aux moyens de compléter les resultats de la grande expédition antarctique, et d'en augmenter la valeur. Votre Excellence choisèra dans sa sagesse, ce qui, parmi tant d'objets importans pour l'art nantique et pour la Géogruphie physique, pourra lui paroître d'une exécution facile. Je sais borner mes espérances. Je supplie V. E. de jetter les yeux sur quelques additions anx instructions scientifiques que j'ose Lui adresser. C'est presque être présomptucux que de vouloir ajouter à un excellent travail, rédigé en partie par Sir J. Herschel. J'ai cédé aux instances amicales de MM. Sabine et Lloyd, et je vous supplic, Milord, de vouloir bien faire mettre entre les mains de Sir J. Herschel un écrit fragihentaire dans lequel ce grand astronome et savant physicien effacera, librement tout ce qui lui paroîtra peu exact, ou moins digne de l'attention des voyageurs. Je suis. &c. &c. A. DE HUMBOLDT. Additions fragmentaires aux " Instructions for the Scientific Expe. dition to the Antarctic Regions." Les personnes qui sont chargées des observations scientifiques, ayant des connoissances très-variées, il suffit de leur rappeler avec la plus grande concision les points qui paroissant de quelque importance dans le cours de leurs travaux. I. FORME DE LA SURFACE CONTINENTALE. Mouvemens des terres par soulèvement ou par dépression, soit lents et progressifs, soit brusques et instantanés, toujours comme effet de la réaction de l'intérieur fluide d'une planète vers sa croûte plus ou moins consolidée. Il seroit important de placer des marques sur les côtes des continens et des iles, à une élévation rigoureusement déterminée au ·96 dessus des plus hautes marées. Je préfércrois des barres de cuivre préparées d'avance en Angleterre, ayant une inscription de la date et le nom du Capitaine Ross. Un trait ou sillon creusé dans le rocher, réuniroit deux marques métalliques éloignées l'une de l'autre au moins de 15 pieds. Le sillon doit être très-exactement lorizontal. Pour chaque endroit on aura vérifié l'élévation moyenne des marées l'une manière approximativc. Des marques semblables, mais de fer et de 2 pieds de long, ont été placées (à ma prière, après mon retour de Sibérie) par M. Lenz, membre de l'Académie de St. Petersbourg, sur les côtes rocheuses de la Mer Caspienne près de Bakou (voyez Poggendorf, Annalen, t. xxvi. p. 364.). Les barres sont scellées par du plomb fondu. II. MAGNÉTISME TERRESTRE. Tout ce qui a rapport à l'importance de la simultanéité des déterminations d'inclinaison, de déclinaison et d'intensité magnètiques a été consigné dans la lettre que j'ai eu l'honneur d'adresser à Nİ. le Comte de Minto, Premier Lord de l'Amirauté. J'ai rappelé ce qui cst relatif à la forme, et aux directions actuelles de l'équateur nagnétique (courbe d'inclinaison zéro) et les lignes sans déclinaison. Je n'ajoute ici que le désir que l'on puisse observer en outre des époques prescrites par M. Gauss, aux époques astronomiqnement importantes des solstices et des équinones, comme je l'ai fait conjointement avec M.Oltmanns en 1806 et 1807, pendant 5 et 6 jours et autant de nuits. A cause de la plus grande précision des instrumens actuels, 24 ou 36 heures suffiroient. Je signale aussi les points suivans:--Examiner les influentes lunaires d'après les indications de M. Kreil, astronome de Milay. aujourd'hui à Prague; faire attention aux corages, aux grandes chutes de grêle ou de neige, aux jours couverts ou sereius ; voir si des changemens atmosphériques modificnt les pihilomènes magnétiques d'une manière sensible et stable : oxuminor si sur por ou sur les glaces polaires, on remarque quelque influence de la constitution minéralogique du fond; si des perturbations locales se font sentir sur mer, là ou l'on peut supposer que les caux ne sont pas trèsprofondes. L'intensité des forces se trouvoit diminuée à la hauteur que M. Gay Lussac a atteinte en ballon : on reconnoit cette diminution, lorsqu'on les observations de ce savant par la température des couches d'air qu'il a parcourues. La position dans un vaisseau à la surface des' mors est une position seniblable; moins par rapport à la surface moyenne de la terre, que par rapport à l'indépendance relative aux attractions locales. Les observations faites sur de hautes montagnes, au dessus de 2500 toises, (observations d'inclinaison et d'intensité recueillies soit par moi, soit tout récemment par d'autres voyageurs) donnent des résultats peu concordans à cause des perturbations dues aux couches soulevées de la croûit icrrestre. Ces considérations sur le décroissement très-lent des forces magnétiques dans le rapport hypsométrique, et sur la petitessc de la prolondeur moyenne de l'océan, méritent l'attention des physicies. Même sur le sol volcanique de Rome, nous n'avons pas trouvé 1.jay Lussac et moi, de différence sensible dans l'intensité de la force horizon- ·· 97 tale au Monte Pincio, à la Villa Borghese, et à Tivoli. Ces expériences seront très-aisées à répéter sur la glace, où l'on peut s'éloigner à de grandes distances du navire, et où les influences du fond de la mer, si elles existent, doivent se manifester au milieu de la marche uniforme des phénomènes d'intensité ou d'inclinaison. Les tremblemens de terre m'ont paru agir quelquefois sur l'inclinaison. Multiplier les observations d'inclinaison horaire là ou les secousses sont fréquentes. Les aurores boréales changent-elles parfois la force horizontale sans influer sur l'inclinaison ? Y a-t-il quelque aspect particulier à cette classe d'aurores boréales ou australes, qui affectent peu les déclinaisons lioraires de l'aiguille? Obse'n vor de priítórence les variations magnétiques aux époques où beaucoup l'étoiles filantes entrent dans l'atinosphère. Examiner si de grandes perturbations (les orages magnétiques) sc répétent pendant plusieurs jours aux mêmes heures ; si en général ces orages magnétiques ne sont pas beaucoup plus fréquens de nuit, lorsqur le soleil 1 règle et ne teinpère plus, par son séjour au dessus cle l'horizon, la marche de l'aiguillc. Il est d'un vif intérêt de découvrir los rapports Ilu magnéti me terrestre (et de ases manifestations variables) avec d'autres phénomènes physiques, soit dans les monvemcir qui ilipendent du tems vrai (du passage du soleil par le méridien de chitar lieu), soit dans les mouvemens isochrones, c. à d. dans call fout vit peut déduire la différence de longitude avec un degré di precision inattendue. III. Mers. Observer les différences de tenpérature dans la haute et la logge mer, comme l'influence que la pente plus ou moins rapide des accores produit sur le refroidissement des bas-fonds; inésurer la distance à laquelle loin banc de glace agissent sur la température des eaux de la surtives. Les températures des couches inférieures de la mer ont acquix it! nouvel intérêt depuis que dans le voyage de Kotzebue, M. Lenz, muni d'excellens instrumens, a trouvé souvent sous les tropiyues (par 17" et par 21° de latitude, à 600 et 900 toises de profonclur) 271 «t 2 41 du therinomètre centigrade, (Poggendorf, t. xx. . 73,) i que l'on sait que le maximuin de densité de l'eau pure n'est pils applicable à l'eau de mer. Le Capitaine Bérard, avec une ligne de sonde d'un millimètre de diamètre, est parvenu à sonder jusqu'à 13:34 toises de profondeur (Bérard, Description rles Côtes d'Alger, 1837, p. 41). Les thermomètres,à minimum et à maximum din M. Magnus et de M. Walferdin sont d'un emploi très-précis amune l'ont prouvé les belles expériences de M. Arago dans les puies watúsiens. C'est dans les courans océaniques d'une haute teupératule quc les sondes thermométriques seroient surtout d'un grami inzi.rêt. Examiner si le courant d'eaux froides qui longe Ics côtes du Pérou jusqu'au Cap Pariña, où il dévie vers les Galapagos, ou dans lequel j'ai trouvé par les 12° de latitude sud l'eau de la surface seulement à 12:4 Réaumur, quand hors du courant, la mer étoit à 22 R., prend sa source 75° à l'ouest du méridien H ·98 du Cap Horn par les 60° et 65° de latitude sud. Ce mouvement des eaux froides est il d'abord dirigé vers le nord-nord-est, et puis (sur le parallèle de 35° sud) vers l'ouest, en frappant contre les côtes du Chili, et se divisant sur ces côtes en deux courans vers le nord et le sud? Examiner la température de ce fleuve pélagique loin du littoral à l'ouest du Chili. (Voyez l'intéressante Carte du mouvement des eaux dans le Grand Océan Austral, par le Capitaine Duperrey, 1831, et l'Atlas Physique de Berghaus, Cahier I. No. 4.) Employer différens moyens, ou de nivellement optique le long des mâts, ou de dépression d'horizon (si le soleil est visible), ou de privation du vent par les vagues (en mesurant en même tems l'inclinaison de la mâture,) pour déterminer, dans une tempête, loin des côtes, le maximum si souvent contesté de la hauteur des vagues. J'ai cru trouver ce maximum de 43 piés français, par le moyen de la dépression de l'horizon, dans la Mer du Sud, pendant une de ces tempêtes désignées sous le nom de Papagayos, à l'ouest des côtes de Guatimala. D'autres voyageurs croyent ce chiffre de bcaucoup trop grand. IV. ATMOSPHÈRE. Si la pression moyenne de l'air au niveau des mers diminue dans l'hémisphère boréal, depuis le parallèle de 550 vers les tropiques ct vers l'équatear, elle paroît, par de certaines longitudes, diminuer aussi, entre les 55° et 68°, ct puis augmenter de nouveau. D'après les recherches de M. Schouw, on trouve, à zéro de température et en appliquant avec M. Poggendorf la correction relative à la pesanteur, pour Lignes Christiansborg 50 Bar, 336:09 La Guayra 10 6.16 Palerme 38 8.00 ? Naples 7.82 Londres 514 7.53 Altona. 531 7.35 Danzig 7.24 Edinburgh 56 6:46 Christiania 60 674 Bergen. 60 6:02 Reikiavig. 64 3.89 Godthaab 64 3.86 Upermavik 73 5.49 L'Ile Melville 745 6:35 Spitzberg 75% 6.23 En faisant abstraction de la correction relative à la pesanteur, on auroit, pour La Guayra 336.98, pour Londres 337-38, pour Reikiavig 333.36, pour l'Ile Melville 33561. (Poggendorf, t. xxvi. p. 241 et 475.) Il est important d'avoir ces chiffres sous les yeux pour les comparer avec les hauteurs moyennes du baromètre que * Ancienne mesure de France, Nord. . . . . + 549 . . . . . . . .

·99 les voyageurs obtiendront, réduites à zéro de température, par les différentes latitudes et longitudes de l'hemisphèrc austral. On comparera aussi la direction moyenne des vents de l'année ct des saisons avec la pression atmosphérique. Quant à la direction du tournoiement du vent dans les deux héinisphères, cffct de la rotation du Globe et de la vitesse des molécules d'air correspondant à chaque parallèle, je recommande les ingénicuses recherches de M. Dove dans les Meteorologische Untersuchungen, 1837, pp. 124--138. Déjà Bacon de Verulam a dit, dans le chapitre De successione ventoruin: "Si ventus se inutet conforniiter ad motum solis, non revertitur plerumque." La direction du tournoiement est opposée dans les deux hémisphères (Churucca, Viage al Magellanes, 1793, p. 15); mais ce fait, connu depuis long-tems des marins, n'avoit pas été examine sous le rapport de ces importantes influences météorologiques. Examiner la températures des plages et la comparer à la température de l'air. Placer des thermomètres dans le sol à différentes profondeurs, sous différentes latitudes. L'observation des réfractions par un froid très-intense deviendroit surtout importante, si par l'ascension à quelque montagne voisinc et d'unc élévation considérable, on pouvoit déterminer en même-tems le décroissement du calorique. Dans le voyage de la corvette La Recherche au Spitzberg et aux côtes de Laponie, on s'est servi de ballons captits et de thermomètrographes, mais des expériences de ce genre sont d'une exécution peu facile. Le décroisseinent de la chaleur est si lent pendant de grands abaissemens du thermomètre, que les obscrvations de réfraction de M. Svanberg faites en Laponie par 29° centésimaux audessous du point de la congélation donnent, d'après les forníules de M. Laplace, un décroissement de 243 mètres par degré centésimal. L'Ephéméride des étoiles filantes, publiée par M.Quetelet, sera indispensable aux voyageurs, pour tiser leur attention sur d'autres jours que ceux d'Août et de Novembre. Examiner si les mois de Fevrier et de Mai se signalent particulièrenient, et si les jours des grandes chutes d'étoiles filantes il y a simultanément des traces d'aurores australes. L'Amiral Wrangel assure avoir observé souvent, dans son expédition aux iles des Ours et de Kolioutchin dans la Mer glaciale, que pendant les aurores boréales, certaines régions du ciel ne sont devenues lumineuses que lorsque des étoiles filantes les ont traversées. Les aurores australes laissent-clles des traces pendant le jour : ces traces se manifestent-elles par une certaine disposition linéaire de nuages, (cumulito-stratus du journal de M. le Capitaine Fitz-Roy,) également espaces ? Ces rangées de petits nuages ou bandes polaires m'ont paru le plus souvent dirigées dans le méridien magnétique, p.e. sur le plateau de Mexico, quelquefois, surtout dans le nord de l'Asie, je les ai vues tourner progressivement pendant des nuits très-calmes du N. par l'E. N. È. à l'Est. Ces pparences ne seroient-elles que des effets de perspective, des convergences de stries parallèles, causées par des vents supérieurs - Cette explication me paroît douteuse. H 2 ·100 V. ZOOLOGIE. Les mémorables travaux de mon ami et compagnon de voyage, M. Ehrenberg, et les rapports qu'offreut ces travaux, soit avec la connoissance intime de l'organisation d'animaux que l'on croyoit jadis d'une structure infiniment simple, soit avec la phosphorescence de l'océan, et les grandes questions de la Géologie moderne, invitent les naturalistes voyageurs à diriger leurs recherches sur les points suivans : Recueillir de l'eau de mer partout où l'on aperçoit à la surface des changemens de couleur et de densité, cn forine de pellicules, de stries et de taches huileuses. Dans ces parages on est sûr de trouver abondance d'animaux microscopiques, et conime il est prouvé par l'expérience que ces petits êtres, même les infusoires pêlagiques, peuvent, soit à cause de leurs carapaces siliceuscs, soit à cause de la consistance de lcurs membranes ou tissus organiques, etre examinés sous le microscope, après avoir été conservés pendant plusieurs années, on tâchera, j'espère, de les recueillir de deux manières. Là où l'on voit les stries de différentes couleurs, l'on enlèvera une portion de ces strics, cn enfonçant dans l'eau de mer des lames très-minces de mica, ou du papier bien fort. Ces lames de mica, ou ces feuilles de papier seront soulevées horizontalement; on les séchera et Ics conservera dans un livre, les animaux restant attachés aux niêines lames de mica ou feuilles de papier, au inoyen desquelles on les a recueillis. Dans les parages où l'eau de la mer paroît entièrement pure et presque incolore, cllc renferme souvent des acalèphes, des crustacées, et des infusoires microscopiques. Il s'agit seulement de pouvoir Axarpiner ces eaux dans un état pour ainsi dire concentré, d'enferiner les êtres vivans dans un moindre volume du fluide. À cet effet, M. Ehrenberg a l'habitude de faire puiser de l'eau à la surface de la iner dans un seau, et de la faire tiltrer on passer à travers un linge très fin, de manière que chaque fois on sépare la portion conceitrée qui n'a pas encore pénetré entièrement à travers le filtre. Ces portions concentrées (plus riches en animaux invisibles à l'ail nu) sont conservées dans de petits flacons de 2 ou 3 pouces de haut. On laisse quelques bulles d'air entre le bouchon et l'eau. Si pendant l'opération de la filtration, on découvre quelques acalèphes visibles à l'oeil nu, il faudra les séparer et les placer dans l'esprit de vin, afin que ces petites anasses gélatineuses n'altèrent pas l'cau de mer concentrée. Comme ces opérations sont extrêmement faciles, il est à désirer que pendant toute l'expédition, par différens degrés de latitude et de longitude, on recueille de l'eau de mer surtout là où les algues marines abondent. Par les moyens qu'on vient de proposer, on parviendroit à étendre d'une manière inattendue la connaissance de la nature intime des petites organisations et de leur distribution géographique sur le globe. Il ne faut que les estomacs remplis de carmin et d'indigo dans tous les Polygastres, que les yeux dans les Rotifères et l'Eudorina elegans, et les dents des Hydatines, se conservent pendant de longues années, lorsque ces êtres pas oublier ·101 microscopiques sont préparés entre des lames de mica d'après la méthode de M. Ehrenberg. Recueillir les animaux qui causent la phosphorescence de l'océan, phosphorescence dans laquelle le Mammarid scintillans semble jouer le rôle principal, avec d'autres acalèphes et des infusoires pélagiques (especes de Peridinium, de Synchæta et dc Prorocentrum). Observer si la phosphorescence n'est pas plus générale et plus fréquente par ạn ciel couvert et d'apparence orageuse : examiner si pendant la phosphorescence générale de l'Océan l'Æquorea forskoeliana, A. phosphorifera, le Pelagia cyanella, P. noctiluca et P. pancpyra, luisent par scintillation c. à. d. non d'une manière continue, mais en donnant des étincelles par la décharge spontanée de certains organes électriques cellulaires; conserver ces acalèphes et les béroés phosphoriques dans l'esprit de vin ; examiner si certaines espèces de poissons, Chimæra arctica, Clupca erythræa, Coryphæna hippuris ct Scomber pelamys, sont phosphorescentes par elles-mêmes, ou si ce n'est pas plutôt par l'adhérence d'infusoires phosphorescens. Recueillir partout celles des substances minérales qui selon les découvertes de M. Ehrenberg, sont composées d'infusoires fossiles comme le tripoli, les schistes à polir, les semi-opales, les minérais de for limoneux riches en Galionella ferruginea, les dépôts colorés des sources salées ou ferrugineuses, les craies, les silex Pyromaques, les marnes alternant avec la craie, les dépôts siliceux, les terres que mangent quelques peuples, par goût ou par besoin. Quoique les infusoires à carapaces siliceuses, plus iudifférens aux variations des latitudes et des clinats, ayent manifesté généralement plus d'aptitude à résister aux grands cataclysmes géologiques, plusieurs des pryozoa calcaires ou polythalames de la craie existent cependant aussi vivans dans la mer actuelle. M. Ehrenberg en a récemment trcanvédlo vivans dans la mer Baltique, identiques avec des polythalames enfouies dans les craies et leurs marnes. Cette circonstancc d'identité clonne, à cause de l'âge des formations craicusrs, un vif intérêt géologique à ce genre d'investigations. Recueillir et conserver avec soin, à cause des petites organisations qu'ils renferment, les sables des dunes, les sables de toutes les côtes que l'on visite, les sables rejetés par les hautes marées, les sables qui s'attachent à la sonde et à l'ancre des vaisseaux; recueillir des échantillons des terres qui composent les marais et des endroits inondés et desséchés. Les plus petites quantités suffisent, en marquant bien exactement les localités où elles ont été receuillies. VI. BOTANIQUE. Plantes marines qui vivent en société. Il reste des doutes, si dans certains parages (comme au banc d'algues anciennement connu près des Iles Azores) le Fucus natans (Sargassum vulgare et S. bacciferum, Agardh) continue à végéter sans racines, en flottant à la surface de l'Océan au gré des vents et des courans, ou si le Fucus, récemment arraché à des rochers dont on suppose l'existence et la proximité, ne peut conserver son état de ·102 fraicheur que pendant un très court espace de tems. Les ingénieuses considérations de M. Charles Darwin (Journal, pp. 303-305), ont répandu un nouvel intérêt sur ces great aquatic forests." L'analogie des Vaucheries et du Polysperma glomerata, la facilité même avec laquelle, dans l'eau douce, des plantes phanérogames (l'Aldrovanda vesiculosa, et des branches du Najas major) continuent à végéter, lorsqu'elles nagent dépourvues de racines, ont fait croire à un voyageur d'une instruction très-variće, M. Meyen, que le Fucus natans peut pousser des feuilles (frondes), sans racines et sans adhérence au fond, mais que dans ce cas, le Fucus natans flottant ne porte jamais de fruits. Recueillir les échantillons de Fucus qui se sont développés en forme arrondie, les branches s'étendant comme par rayons. Mesurer la teinpérature de l'eau la plus froide dans laquelle végètent ces plantes sociales. (Signí) ALEXANDRE DE HUMBOLDT. Berlin, le 26 Octobre, 1839. 6 Letter from Professor A. Erman to Major Sabine, R.A., F.R.S. MONSIEUR, A Berlin, le 12 Nov. 1839. J'ai eu le plaisir de vous exprimer à Berlin mon vif intérêt pour l'expédition magnétique, dont votre rapport sur l'intensité totale (Seventh Report of the British Association for the Advancement of Science) a fait concevoir le plan, et que le gouvernement Anglais a mis en oeuvre avec une munificence entièrement digne du sujét. Vous savez combien je félicite: les voyageurs qui continucront jusqu'aux plus hautes latitudes Australes, les observations que je n'ai poussées que peu au delà du parallèle du Cap Horn, et qui vont tout autant préciser la forme et la position des lignes magnétiques de cet hémisphère, que nous avons pu le faire, M. Hansteen et moi, pour celles de l'Asie du Nord où tout conspiroit à favoriser notre entreprise. Aussi est ce avec beaucoup de reconnoissance que j'accepte l'entremise que vous avez bien voulu m'offrir, pour signaler aux membres de cette grande expédition, quelques résultats et quelques sujets de recherche, que notre voyage pour un but analogue me porte à recommander à nos successeurs. Il est vrai que la belle instruction dont la Société Royale a muni ses voyageurs, leur indique très complètement les moyens d'obtenir, tant sur mer que lors des mouillages aux côtes et aux glaces polaires, une série continue de déterminations des trois élémeus magnétiques. Elle est si riche en détails importans que, dans l'intérèt de la science, on ne sauroit rien desirer au-delà du stricte accomplissement de ce plan de voyage. Cependant pour toujours soutenir l'attention et le zèle dans un travail uniforme d'aussi longue durée, et pour les faire redoubler à point ·103 nommé, dans les endroits où les observations augmentent d'importance, il n'y a je crois rien de plus efficace, qu'une comparaison suivie des résultats de l'observation, d'une part avec ceux de la théorie qu'il s'agit de perfectionner, et de l'autre avec les évaluations purement empiriques de ses prédécesseurs. L'expédition antarctique doit iouir de cet avantage pour ses mesures d'intensité totale, en se servant de la carte isodynamique construite d'après la théorie de M. Gauss, et de celle que vous avez directement établie sur les résultats des voyageurs*. J'ai destiné aux mêmes fins la carte ci jointe, Pl.III., représentant les lignes d'égale déclinaison pour une époque entre 1827 et 1830. Je les ai obtenues par une interpolation graphique, et devant fournir des isogones indépendantes de toute vue de théorie. J'ai noté sur la carte même les résultats numériques qu'elle représente, et il ne me reste ici qu'à mentionner les voyageurs qui les ont fournis, la direction des routes qu'ils ont suivies, et l'époque de leurs observations. I. L'Europe et l'Asie septentrionales. MM. Hansteen et Due, de Christiania à Irkuzk et à l'embouchure du Jenisci, en 1828 et 29. Erman, de Berlin aux bouches de l'Obi, par Irhuzk et Ochozk au Kamtschatka, en 1828 et 29. II. Le Grand Océan. Lc Capit. Lütke (sur la Corvette le Siniavine), du Cap Horn, par Valparaiso, les iles de Sitka et d'Ounalaska, à Petropawlowsk, en 1827. Idem (idem) de Petropawlowsk à Manilla, en 1828. Erman (la Corvette le Krotkoi), de Petropawlowsk, per Sitka, San Francisco, et Otaheite, au Cap Horn, en 1829 et 30. III. L'Atlantique Le Capit. Lütke (le Siniavine) de l'ile de Teneriffe, par Rio Janeiro, au Cap Horn, en Décembre 1826, et 1827. Idem (idem) du Cap de Bonne Espérance, par les iles de St. Helène et de Fayal, au Canal Anglais, en 1829. Erman (le Krotkoi), du Cap Horn, par Rio Janeiro, à Portsmouth, en 1830. IV. La mer des Indes. Le Capit. Hagemeister, (la corvette le Krotkoi), du Cap de Bonne Espérance, à Port Jackson, en 1828. Lütke (le Siniavine), de Manilla, au Cap de Bonne Espérance, en 1829. Je n'ai du ajouter à ces résultats presque contemporains (Dec. 1826 à Oct. 1830) qu'une disaine d'observations antérieures, toutes faites dans la mer glaciale du Nord, et nommément par le Capitaine Wrangel dans la partie orientale de cette mer (68° à 70° lat., 162° à 1820 à l'Est. de Green.) en 1823, et par # Pl. I. and II.

·104 $ 1. Le Capitaine Lütke dans sa partie occidentale (70° à 770 lat., 27° à 52° à l'Est de Green.) en 1821. . Si l'on compare maintenant dans leur ensemble ce tracé immédiatement calqué sur les observations, et la carte que la théorie de M. Gauss a fournie pour la même époque, Pl. IV., on sera frappé de leur accord éminemment satisfaisant, tant pour les formes que pour les places qu'elles assignent à la plupart des isogones. On envisagera toutefois, comme prévues d'avance, des courbures plus accidentées et moins arrondies dans les isogones empiriques : résultats nécessaires tant d'une interpolation imparfaite d'observations affectées d'erreurs, que d'influences locales, telles que la différente constitution géologique des pays et leurs accidens de climat; car la théorie que son illustre auteur ne présente que comme une ébauche, ne saurait déja reproduire ces effets de causes secondaires. Mais, indépendamment de ces écarts accidentels et locaux, une comparaison suivic des deux cartes fait ressortir entre elles quelques différences plus décidément prononcées, portant sur de grandes portions d'isogones bien établies par l'observation. Je me permets de les signaler ci-après à l'attention de vos voyageurs. Entre (jo et 1500 E. 1. Les sommets concaves des isogones negatives (orientales) que la carte empirique place vers 770 E., y atteignent de moindres latitudes que d'après la théorie. Nommément: Sur la carte empirique. Sur la carte de M. Gauss. L'isogone de -15º descend jusqu'à 659 lat. 78° lat. 100 58° 5' 640 2. Le système de déclinaison positive ou occidentale, qui a son centre d'après l'interpolation graphique rers 130° E., et d'après M. Gauss àpcine (u•3 à Touest de ce même méridien, s'écarte de la théorie par la valeur des lignes qui le composent, et cette différence est l'inverse de la précédente. Les sommets convexes de ces lignes sont, Sur la carte empirique. Sur la carte de M. Gauss. Pour la courbe de 0° à 680.5 lat. 610.7 lat. + 204 65° 51.0 + 6° vers619 n'existe pas, le centro du système situé en 45º lat. ne devant avoir que + 20 30' déclin. On résumera ces deux circonstances, en observant qu'un voyage depuis 650 lat. et 77º E. long. jusqu'en 61° lat. et 130" E. long. offre en réalité un plus fort changement de déclinaison que suivant la théorie. En effet le changement observé serait de 21° (depuis -15° jusqu'à + 6°) où la théorie ne demande que 10° (depuis 10° jusqu'à 0°). 3. La différence des deux cartes relative à la ligne sans déclinaison entre les dits méridiens n'est au fond qu'une suite de ces deux circonstances (Nos. 1 et 2). La branche occidentale de cette courbe, " > ·105 sur laquelle la théorie et l'observation sont presque d'accord, et que cette dernière fait passer par 50° lat. et 48° E. long., se distingue sur les deux cartes par son prolongement vers le Sud et le Sud Est. Elle a son sommet concave sur la carte empirique en - 10 lat. et sur la carte de M. Gauss vers -10°. 2 lat. (les latitudes australes étant pris négatives), et passé ce terme, d'après l'observation directe, la courbe se relève vers le Nord Est et le Nord, embrasse le système Asiatique de déclinaison occidentale, pour ne se replier qu'après, par la mer d'Ochozk, le Grand Océan, et la mer des Indes, sur la nouvelle Hollande. La théorie lui assigne au contraire, d'abord après le sommet concave, un rebroussement vers le Sud (en 105° E.), qui la porte directement sur la nouvelle Hollande; aussi voit on sur la carte de M. Gauss, le susdit système de déclinaison occidentale, entouré d'une courbe à zéro fermée et isolée, et dont la branche orientale se trouve plus à l'ouest que ne le demandent les observations pour la partie correspondante de la ligne continue. En effèt ces parties correspondantes de la coutbe à zéro coupent : Sur la carte empirique. Sur la carte de M. Gauss. Le 60° lat. sous 150° E., sous 1400 E. 500 1519:3 E. 1470.8 E. Mais je suis loin d'attribuer une spécialité d'intérèt à l'isogone de zéro;- la différence de deux branches isolées, à une courbe continue, que nous venons de lui trouver sur les deux cartes, ne me parait au contraire ni plus ni inoins grave, que si elle portait sur quelque autre courbc de ce genre. Je crois plutot que pour voir cet écart dans son vrai jour, il faudra observer que dails le systéme en question, la valeur limite entre des courbes isolées et des courbes continues cst, suivant la théorie de 1° 14' déclin. Drithtalc, tandis que l'observation parait l'élever à une déclin. occidentale de + 1", ou environ, car pour dériver de données numériques la valeur précisc de cette limite, il ne suffit ni d'une interpolation graphique ni d'aucun moyen différent d'une théorie complète. J'observe en outre, et pour me prémunir contre plus de responsabilité que je ne dois avoir sur ce point, que la dite partie de ma carte repose uniquement sur les observations suivantes de M. Lütke. Longit. Latitude. Déclinaisons d'après : M. Lutke. La théoric. L-5 135° 34' 134 38 134 04 122 33 117 53 115 21 113 03 105 04 105 32 + 14° 35' + 15 34 +16 4 + 19 54 +13 41 13 4 + 12 41 8 39 9 46 L. +0° 10' +04 +1 2 + 2 20 + O 33 + 0 23 +0 33 +1 26 +10 T. - 10.25 -1 0 +0.2 - 0.8 -0.8 1.4 + 10.4 +1:1 +1.0 +2:1 +1:3 +1.2 + 2.0 +1.8 +1.2 - 02 ·106 Malgré leur grande influence sur la forme de la courbe à zéro, ces différences entre la théorie et l'observation sont donc beaucoup plus foibles que celles observées dans les contrées précitées (Nos. 1 et 2) et qui s'élevaient respectivement à -5° et à + 6º. Les parties australes des deux cartes situées entre les dits méridiens de 0° à 150° E, s'accordent très bien entre elles. § 2. Depuis 150° jusqu'à 360° E. Il en est de même dans l'hemisphère boréal, depuis 162° jusqu'à 262° E., pour les isogones de -30°, à -15°; mais passé ce terme les courbes théoriques de -12°, de -10°,etc. portent les déclinaisons orientales qu'elles expriment jusqu'à de moindres latitudes boréales que ne l'indique l'observation. Ainsi Sur la carte empirique. La carte de M. Gauss. L'isogone de - 12° descend jusqu'à 330-5 lat. 230 lat. 60 28°.5 17° C'est cette circonstance, et une toute pareille pour les isogones de même nom dans l'hemisphère austral, qui produisent 4. Une diversité des deyx carics relativement au système fermé de déclinaison orientale dans le Grand-Océan. Les courbes de - 10°, -9°, et -8° sont les parties de ce système que l'observation directe a le mieux reconnues, et nous trouvons à chacune d'elles plus d'étendue dans le sens du méridien que ne l'adopte la théorie. Ainsi d'après La carte empirique. La carte de M. Gauss. ſ depuis + 270.8 lat. + 170 lat. L'ivog: de 10° va jusqu'à - 4.90.5 - 39 ſ depuis + 25 + 120.5 go Ljusqu'à - 47 36 ſ depuis + 23 + 70.5 go ljusqu'à - 44005 340 Leurs diamètres dans le sens du méridien sont donc respectivement : d'après Tobservation directe de 7703, 72° et 670-5, et d'après l'interpolation théorique do 56°, 43º.5 et 410.5. Les observations nous apprennent en outre sur les isogones de ce système, que celle de-109 passe bien décidément d'un pole nord à un pole sud, et qu'au contraire la courbe de - 89 est isolée et rentrante. L'isogone de 90 participe tellement aux propriétés de ces deux espèces de courbes, que, parmi celles que représente ma carte, elle doit être la plus voisine de la valeur limite. La théorie s'y accorde très bien en indiquant -8° 46':5 pour cette même limite. L'accord des deux cartes est moins parfait sur la position du centre de ce système et sur la déclinaison qui y règne, car la forme des courbes enpiriques de-10" à -7° engagerait à le présumer situé vers 2310.8 E. -120 lat. 9 - ·107 . tandis que l'interpolation théorique le porte en 2190.8 E. -140.5 lat. aussi la théorie attribue à ce point une déclinaison minimum de -5°15', mais nous avons très souvent observé sur le Krotkoi, de même quc Mr. Lütke sur la route plus orientale du Siniavin, des déclinaisons entre-5°ct-4° et mêmes quelques unes de-3°50'à --3040'. Nous étions cependant encore très sensiblement eloignés du centre des courbes. 5. Dans l'hemisphère austral entre les méridiens de 192º E. et 262° E. les isogones de-12° et de-15°, d'après la théorie ne s'approcheraient du pole austral que jusqu'aux paralleles de -38°.8 et de - 49º, tandis que les observations paraissent les y étendre jusqu'en --520.7 et - 58º. Mais je termine cette comparaison, en vous priant, Monsieur, d'accorder encore votre attention à l'harmonie très parfaite des deux cartes relativement au système de déclinaison occidentale qui re. couvre l'Atlantique et l'Europe, lequel, vu la grande fréquence des observations dans ces parages, est un des plus solidement établis par l'expérience directe. La théorie porte la raleur limnite pour ce système à +22° 13', et les observations démontrent, d'abord, que cette même valeur est comprise entre +20° et +25°; et de plus, qu'elle est beaucoup plus rapprochée de la moyenne arithmétique de ces doux nombres, que d'aucun d'eux. Je passe à quelques autres sujets de recherche que je me permets de résumer en de simples questions : Aurores Polaires. Y a-t-il dans les hautes latitudes antarctiques quelque indication d'une duplicité d'Aurores Polaires ? Les habitarr de Bérésow sur l'Obi (630 56' N. lat.) en distinguent en effet de deux espèces: de plus loibles ayant leurs centres à l'Ouest; et de plus brillantes qui se forment à l'Est du meridici. Les données que j'ai pu recueillir sur les azimuts où se formeraient ces deux phénomènes m'ont fait présumer qu'ils surgissent respectivement de l'un ou de l'autre des deux foyers magnétiques de l'hemisphère boréal. Phosphorescence de la mer. Existe-t-il un minimum de température qui empècheroit d'avoir lieu le développement de lumière opéré par des Méduscs, des Scolopendres, et surtout par des Crustacés microscopiques vaguement qualifiés de Zoophytes infusoires ? J'ai observé des phases très brillantes de ce phénomène sur la mer d'Ochozk dans les nuits du 3 Aout par +3902 F. température de la mer. 5 Aout +41 .0 6 Aout +41 2 7 Aout +46.6 mais il y eut cessation parfaite la nuit du 4 Aout où la température de la mer baissa jusqu'à +3704 F. Un phénomène lumincux, différent d'apparence et d'origine, s'observe sur la même mèr d'Ochozk à l'approche de la saison froide, et y est connu sous le nom de Spolochi. Ce sont des bandes lumineuses, > >> ·108 qui paraissent douées d'un mouvement propre, et qui se repandent en jets rayonnans, depuis un centre commun à plusieurs d'entr'elles et également situé à la surface de la mer. On ne saurait les confondre avec la lumière due aux Mollusques, laquelle ne forme jamais des bandes, mais des points distinctement isolés, et brillans de préférence dans le sillage du vaisseau, dans les brisans, et sur l'écume des lames et des vagues. Les bandes en question ressemblent au contraire à de la lumière électrique qui se répandrait à la surface d'un conducteur imparfait. Ya-t-il des phénomènes analogues dans l'hemisphère austral, et quelles circonstances paroissent les provoqtuer ? Météorologie. A égalité de latitude, la moyenne hauteur du baromètre m'a paru moindre sur le grand Océan que sur la mer Atlantique : la même différence s'observe-t-elle encore sur les prolongemens des méridiens en question jusqu'à de hautes latitudes Antarctiques ? Sur le grand Océan, entre le 50° et 60° de latitude nord, c'est bien décidément le vent d'Ouest qui augmente la pression atmosphérique au dessus de la valeur moyenne. Cet effet du vent d'Ouest est le contraire de ce que l'on observe en Europe, et il me parait fort intéressant de recueillir sur ce point des données correspondantes dans l'hémisphère austral. J'ai vu avec beaucoup de plaisir que la Société Royale dirige l'attention de ses voyageurs sur la différence de la colonne barométrique aux limites polaires et aux limites équatoriales des zones alizées. Voici les résultats que j'ai obtenus à cet égard de mes propres obscrvations. (Poggendorf, An. der Physik für 1831.) Mois. Longit. Alizé N.E. Mois. Longit. Alizć S.E. Diff, des hant Diff, des haut barom. barom, Mai et Juin.. 167° 3"-154 Mai 179° 211.245 Janvier 233 0.929 Jan. et Fev. 222 2 714 Juil.et Août.. 327 3 .016 Juin et Juil. 324 2 453 Novembre 342 2 668 Deccmbre. . 333 1 4.92 Les différences sont exprimées en lignes du pied de Paris, ct sc rapportent aux colonnes barométriques observécs à la limite polaire et dela liinite équatoriale de la zone alizéc. Essuie-t-on jamais des orages sur mer quand on est à la fois à une grande distance de toute cóte, et loin des limites des alizés ? L'expérience d'une seule année ne m'en a fourni aucun exemple. Etoiles filantes. Si les voyageurs antarctiques ont le loisir d'observer le retour périodique des étoiles filantes du 10 Août ou du 13 Nov., il seroit fort à desirer qu'ils fissent attention à la direction apparente de leur mouvement, et que nommément ils dessinassent sur une carte céleste leurs trajectoires apparentes, pour examiner après, si leurs ·109 > prolongemens se coupent en un même point du ciel. Les étoiles tombantes de 1839, Août 10, convergeaient toutes vers le point du ciel: Ascens. droite 2220-4 Declinaison 519.2 Australe. L'instruction de la Société Royale ayant dirigé l'attention des voyageurs sur la Chionis Alba, je me permets de leur communiquer que nous avons pris dans deux jours consécutifs 5 ou 6 de ces oiseaux, en nous trouvant à la hauteur des iles Falklands, et à une distance de 48 à 60 milles marins de leur extrémité orientale. Il est vrai que ce fut par un vent d'ouest assez frais, mais on ne saurait cependant décider sans recherche ultérieure, si ces oiscaux intéressans ne sont pas engagés par quelque autre cause à s'éloigner des côtes qu'ils habitent. Les Chionis nous ont paru si propres à la domesticité que l'on pourrait sans doute cs conserver vivantes jusqu'au retour en Europe, et peut être même en faire propager l'espèce. Il est fort probable qu'ils se nourrissent de préférence de grains ou d'autres semences dures, car nous lcur avons constamincnt trouvé dans l'estomac du sable et de petits cailloux, que nous croyidns destinés à faciliter la trituration et la digestion de pareitles substances. Je termine ces remarques detachées en me recommendant, Monsieur, à l'indulgence que vous daignates si souvent témoigner à Votre très devoué serviteur, M. le Major Sabine, R.A. A. ERMAN.

·Comparisons made at the Royal Society's Apartments, Somerset-House, on the 16th, 17th, 19th and 20th of August, 1839, with the Royal Society's Standard, of the Marine, Standard, and Mountain Barometers supplied to the Antarctic Expedition, and to the four fixed Magnetic Observatories of Canada, Cape of Good-Hope, St. Helena, and Van Diemen's Land: by J. D. Roberton, Assist. Sec. Royal Society. Royal Society's Standard. Marine Baroin., No. 19. M. Marine Barom., No. 20. M. Standard, No. 31. Standard, No. 32. Standard, No: 33. Standard, No. 34. Att. Barom. Crown Glass. Barom. Flint Glass. Att. Ther. Att. Ther. Barom. Barom. Att. Ther. Barom. Att. Ther. Barom. Att. Ther. Barom. Barom. Att. Ther. Ther. 30-014 29.930 29-696 29.654 29.634 29.670 29.712 30-080 30-006 29.926 29-690 29-650 29-630 29-666 29.704 30-072 70.2 66.7 62.7 71.2 670 64.5 66.2 60-2 29.972 29.902 29.676 29-614 29.610 29.628 29.676 30-020 62.4 62.7 61.7 62.4 63.3 62.3 63.2 60-4 29.986 29.90€ 29.686 29-616 29-618 29.618 29-684 30-010 62.7 65.2 62-3 62.9 63.4 62-6 633 60-7 30.006 29.918 29.688 29-642 29.632 29-661 29-708 30.066 66.2 66.0 63.7 664 60.8 65.0 66.2 60.9 :30:000 29.916 29.686 29.634 29.620 29.662 29.700 30.068 62.8 62.8 62.0 62.7 64.3 62.6 63.2 60-6 30-010 29.918 29.688 29.642 29.632 29-664 29.708 30-066 63.2 63.2 62.0 63.2 63.7 62.7 63.3 60-7 30.004 29.918 29.686 29-610 29.628 29.666 29.712 30.078 65-3 65.7 63.6 65.3 66.7 64.8 65.8 60-6 Comparisons of Mountain Barometers corresponding with the above Standards. Mountain, Mountain, Mountain, Mountain, 32 No. 31. No. No. 33. No. 34. 1 29.972 29.392 29-660 29.614 29-598 29.638 29-674 30-028 65.6 66-2 64.0 65.7 66.8 65.0 66.3 60.9 29.974 29.888 29.664 29-622 29.612 29-642 29.686 30-032 65.8 66.0 61.2 65.8 66.8 65.3 66-5 61.0 29.998 29.920 29.674 29.622 29.627 29-652 29.690 30-040 62.3 62.5 61-6 62.3 62.8 62.2 62.7 60-2 29.990 29.908 29.692 29-636 29.620 29-656 29.696 30.048 65-3 65.7 64.8 05.4 66.6 64.9 66-0 60-9 ·111 Comparisons made at the Royal Society's Apartments, Somerset- House, in August 1839. Royal Society's Standard Baronieter. Major Sabine's, (Round). Major Sabine's, (Flat). 1 Flint Glass. Crown Glass. Att. Ther. Barom, Att. 1 her. Barom, Att. Ther. 29.634 29.670 29-712 30.080 29.630 29-666 29.701 30.072 67.0 64.5 66.2 60-2 29.578 29.616 29-662 30.012 66.8 65.4 66.5 613 29.590 29.616 29-654 30.006 66.8 65.0 66-3 60.9 Captain J. C. Ross's Instrunyents. Standard Barometer, No. 32. Mountain Barom., No. 32. (1.) Mountain Barom., No. 32. (2.) . 29.020 29.062 29-700 30-068 64:3 02-6 63.2 60-6 29-612 29-642 29.686 30032 66.8 ! 29-612 65.3 : 29.642 66.5 29676 61.6 30.028 63.3 62:3 63.2 60:3 Coniparisons inade in March, 1840. Royal Society's Standard. Standard, No. 46. Mountain, Mountain, Mountain, No. 41. No. 45. No.46. Capt. Koss. Capt. Ross. Lt. Wilınot. Mountain, No. 47. Barom, 1 Barom, Flint Crown Gla.s. Glass. Att. Ther. Barom. Att. Thor. i Att. Barom, Ther, Att. Burom. Ther. Barom. Att. Ther." Barom. Att. Tber. ! 29-661 29.656 -16-3' 29-678 41.3 29.046, 462 | 29.632 47.4 29-690 46-8' 29-062 46-4 29.720 29.718 48.5 29.728 46-229-676 403 29:672: 1704 24.708 169 29.684 466 30.024 30-016' 47.9 30-028 49:6; 29.902 490) 29.952, 50-2! 29.986 19:329-966 19-2 30.266 30.258 456 30-268 43.8 Comparisons of Standards with the Royal Society's, made for the Honourable the East India Company, in April 1840., Royal Society's Standard. Stundard, No. 48. Standard, No. 49. Barom. Barom. Flint Glass. Crown Glass. Att. Ther. Barom. Att. Ther. Barom. Att. T'her. 30-408 30-368 30-346 30.258 30.134 29.918 30-400 30-360 30-338 30-250 30.128 29.912 48.2 4708 48.0 50.0 48.8 49.2 30-126 30:380 30-356 30-274 30-150 29.938 45.7 46.6 47.0 46.2 47.8 48.7 30-400 30-364 30:340 30-250 30.130 29.924 43.9 46.6 47.0 46-3 48.0 48.8 ·Comparisons made with the Royal Society's Standard, of various Standard an! Mountain Barometers, for the Honourable the East India Company, (made in February, 1810.) Royal Society's Standard. Standard, No. 37. Standard, No. 38. Stardard, No. 40. Standard, No. 41. Standard, No. 42. Standard, No. 43. 1 Barom. Flint Glass. Barom. Crown Glass. Att. Ther. Barom, Att. Ther. Baron.. Att. Ther. Barom. Att, Ther. Barom Att. Ther, Barom, Att. Ther, Baroni. Att. Ther. 30-450 30-450 30-458 30-452 30-434 30-442 37.9 30-444 37.6 30-450 3702 30-446 36.6 30-426 36:9 30-458 38.6 30-456 38.2 30-460 38.0 31456 372 30-438 373 30-440 30-412 30-454 30-452 30.128 384 37.9 39:2 37.0 37.3 30-406 38.7 30-418 38-5 30-468 38.5 30-172 373 30-452 3707 30.155 31-4.is 30:168 30--6- 30'444 58.4 37.9 38.0 371 37.3 30-458 30.458 30.462 30.458 30-136 38.0 377 37.5 36.7 37.0 30-456 30-452 30-458 30-452 30-136 38.8 38.4 38.4 37.2 37.6 Printed by Richard and John E. Taylor, Red Lion Court, Fleet Street. Comparisons of Mountain Barometem corresy ding with the above Standards. Mountain, No. 40. Mountain, No. 41. Mountain, No. 42. Mountain, No. 13. Mountain, No. 41. 30-380 30-392 30-390 30-390 30-384 38.6 38.2 37.8 36.9 37-0 30-438 30-132 30.432 30-432 30-420 38.9 38.4 38.2 375 37.6 30-350 30-348 30-316 30-350 30-132 38.8 38:3 38.2 37.2 37-4 30-126 30.432 30-310 30438 30424 38.8 383 384 37-3 37-5 30.424 39.0 11426 38.6 .422 38-3 30.426 3704 30-418 37.6-) ·FORMS FOR BOOKS.

·18 .. duy, the of OBSERVATORY at ..--) DailyObservations of MAGNETOMETERS, DECL MAGVETOM. HORIZ. FORCE MAGVT TOM. MERT. FORCE NAGNETON. Goll Miran TNT Mean at Tor. Observ. Readingi. wenn). Rradinga. in can Thesin. Rradingi. neun. Therm. B .. 1 11 G 2 3 Hille 5 ': Triple Observ! * H 9 I K M in FORN No. 1. * As the days commence with the Magnetic hour nearest after 12 midnight of the place, the time of the triple Observation will be different at pach obser. d. camner therefore printed ·day, the OBSERVATORY at (- and METEOROLOGICAL INSTRUMENTS. 16_ of 18 - Nr1 B.IROVCTER. TURVOY. ELECTROY. Temp. 7 nr Obrero Height. Trmp. l'et. Dry. Point, U More. a d e h. PIL Sluto of the Weather. u Nate 1 d e . 9 L i k 1 LALITEXPER, Date ) Min. / -- > Lau.Ration SOLL de Terrc. FORN No.1.

·day, the of OBSERVATORY at ( Term day Obsery of MAGNETOMETERS, &c. 18_ Times of Observation. MAGNETOMETER. Times of Observation. MAGNETOMETER. Bpecifcation of Magnetometer. Gott. M.T.at M.T. Observ. Rcadings. mean. Gott. M.T. At M.T. Obser. Reading. mcun, Opm 10 par m. .. In. . Declination 0:00 11:30:00 Horizontal Force 2:30 3:30 Declination 5:00 35:00 Vertical Force 1:30 37:30 Declination 10:00 40:00 Horizontal Force 12:30 42:30 Declination 15:001 145:001 Vertical Force 17:30 47:30 Declination 20:00 50:00 Horizontal Force 22:30 52:30 Declination 25:00 55:00 Vertical Farce #27:30 57:30) TIIERMOMETERS. BAROMETER ELECTROM. Hour of Gott, Mean Time, H.F. V.F. Drs. W'et. Helyht. Temper. 10PM il'eatlier. FORM No. 2.

·OBSERVATORY at (
ABSTRACT of daily Observations made during the Month of
18
(NANE OP INSTRUMENT.)
Date.
Sums.
a
6
C
d
e
f
9
h
2
m
Sums
Means!
FORM No. 3.

·OBSERVATORY at -
ABSTRACT of daily Obscrvations made during the Month of..
18
TENP, OF DEN POINT.
ELAST. OP VAPOUR,
TEMPERAT! RADIATION
Dale.
Noon's
Age.
Max. Min.
Solar. Tere.
3 AN.9 AM.3 PN. 9 PM. 3 AN. 9 Am.3 pp. 9 PM.
.:
Meanoli
FORN No. 4.

·)
OBSERVATORY at (
ABSTRACT of triple Obsery, made during the Month of
18_
VFMAGNE CON
DECL MAGNETON HP MAGNETOM
Date
Sums
Means
FORM No. 5.
The Sums and Means of the triple Observations are not required, they have
been included in the form, of which a portion i blank to admit of the inner
fron if necessary. of other observations in the same name

·FORN No. 6.

·day, the OBSERVATORY at ( ABSTRACT of Term-day Observations 18. DECLINATION MAGNETOMETER. Gott Mean Tline. ||10 PM 11. Mida, 1 AM. 2. 3. 4. 5. 6. 7. 18. 9 AM. m. 0:00 5:00 10:00 15:00 20:001 25:00 30:001 35:00 40:00 45:00 50:00 55:00) 10 AM 11. Noon. 1 rm.2. 3. 4. 5. 6. 17. 8. lim. n. 0:00 5:00 10:00 15:00 20:0011 25:00 30:00 35:00: 40:00 45:00, 50:00, 155:00 Form No. 7.

·day, the 07 OBSERVATORY at ( ABSTRACT of Term-day Observations 18) HORIZONTAL FORCE MAGNETOMETER. Gott. Mean Time. 10 PM 11. Midn. 1 AM.2. 3. 4. 5. 16. 7. 8. 9 AM m. . 2.301 12:30 22:30 32:30 -12:30 52:3011 Cherm 10 AM 11. Noop. PM.2. 3. 6. 17. 8. PM 2:3011 12:3011 1:30| 32:30 4:2:30 152:30. Thormit VERTICAL FORCE MAGNETOMETER, 10 PMII. Mida. AM.2. 5. 16. 7. 19 AM 7:301 17:30 27-301 37:30 47:304. 57:301 Theru 10AM 11. sovit. 11 PM.2. lo. 8. PN. 7:3011- 17:30 27:313.. 37:30 17:301 57:30 Therm FORM No. 8.

·OBSERVATORY at (- ABSTRACT of Term-day Observations day, the .18 BAROMETER. THERMOM. ELECINOM. Gott Mcan Tine. STATE OF TIIE WEATIIRR. Height. Temp. Dry. Wet. 10PM 11. Mi.In. LAM. 3 AM. Meaus an. 5. 16. 17. 8. DAN Means 10AM 1). oon 1 pm 62 3.par. Means! 4 PM 5.. 6. 8. 19 PM. Means FORM No. 9.

·Printed by J. L.. Cox and Sons, 75, Orent Queen Street, Lincoln's-Inn-Fields.

·No text on this page.

·No text on this page.

·No text on this page.

·No text on this page.

·C195