Narrative of the surveying voyages of His Majesty's ships Adventure and Beagle, between the years 1826 and 1836

This appears a fitting opportunity to remark, how much the establishment in England of a depository for magnetic needles is needed; whence officers, and persons desirous of making such observations, might be supplied with instruments, which had been kept a sufficient time to have attained their permanent magnetic state, and had been examined from time to time to prove that they had done so. The correction for temperature should be ascertained for each needle, and given with it; as well as the time of vibration (or whatever else constituted the measure of intensity, – as, for example, the angle of deflection in Mr. Lloyd's statical needles,) – observed at the spot which should be selected as most suitable for a point of general comparison; and the observations should be repeated at the same spot on the return of the needle. The want of such an establishment has long been greatly felt; and opportunities, where nothing was wanting but proper instruments, have been lost in consequence, where determinations of great value might have been obtained, in parts of the world of the highest magnetic interest, and where such opportunities are of rare occurrence.

The corrections necessary to render the times of vibration at the different stations strictly comparable with each other, are as follows.

1st, For the rate of the chronometer.

2d, For the temperature of the needle.

3d, For the arc of vibration.

4th, For any change in the magnetic condition of the cylinder.

In extensive voyages, the last-named correction, or that for the change in the cylinder itself, is the one which requires principal consideration. The corrections for temperature, and for the arc, on the first of which particularly much stress has sometimes been laid, are important when extreme accuracy is sought; as for example, in comparing the force at stations which have served as the unities of different observers: their effect is, however, of little moment in observations which include great differences of the terrestrial intensity. But when the magnetic condition of the needle has varied, and interpolation becomes necessary, the instances are rare in which it can be done with entire satisfaction.

The time of vibration of this cylinder at Plymouth, with corrections applied for the chronometer's rate, the temperature, and the arc, was, in December 1831, 770,8 seconds, with a dip of 69° 27′,6, and in October 1836, 777,3 seconds with a dip of 69° 17′,5. The difference in its magnetic state, at the two periods, is shown by the squares of the times of vibration multiplied by the cosine of the dip observed at the respective periods. This makes known what the time of vibration of the same cylinder would have been, had it been free to move in the direction of the dipping-needle, instead of being suspended horizontally; and, consequently, if it had been acted upon by the total magnetic intensity, instead of by the horizontal component only. It is here assumed that the total terrestrial intensity is constant at the same place. This is doubtless not strictly true; but the amount of the change must be too small to require consideration in the period occupied by Captain Fitz-Roy's observations. The horizontal component must, however, necessarily vary with the changes in the dip: and it is, therefore, from the time of vibration in the direction of the dipping-needle, and not from the time of horizontal vibration, that the change, or otherwise, in the magnetism of the needle is to be inferred. We find, then, the equivalent time of vibration of this cylinder in the direction of the dipping-needle to have been 456,4 seconds in December 1831, and 462,2 seconds in October 1836; showing a difference of 5,8 seconds in fifty-eight months. If we compare this amount with the far greater loss of intensity sustained by this cylinder in the preceding voyage, it seems a probable supposition that, at the commencement of Captain Fitz-Roy's voyage, the cylinder had nearly attained its permanent magnetic state; and that its further loss of magnetism, occasioning an increase of 5,8 seconds in the time of vibration, took place in the early part of the voyage: supposing the loss to have been progressive, and not sudden, as from accidents, of which the observations give no indications. Consistently with this supposition, the loss has been distributed through the first half, or twenty-nine months, of this voyage, in the proportion of three-tenths of a second per month in the first ten months, commencing December 1st 1831; two-tenths per month in the next nine months; and one-tenth per month in the remaining nine months. In the last twenty-nine months of the voyage, the intensity of the cylinder is supposed to have been uniform, and the same which it was found to possess on the return to England in 1836.

It is satisfactory that, with this compensation, the observations at Port Praya, in January 1832, and in September 1836, assign almost identically the same relative magnetic intensity to that station.

The correction for temperature for this cylinder not having been previously examined, I received it from Captain Fitz-Roy for that purpose, and made with it the following observations. The cylinder, in its own apparatus, was placed in a large earthen jar, glazed at the top, and standing in a larger earthen vessel, into which warm water could be poured at pleasure, and the cylinder was then vibrated alternately in heated air and in air of the natural temperature. These experiments were made at Tortington, in Sussex.

Whence the formula T′=T [1 × 0.000068 (60°-t)], in which T is the time of vibration at any station, t, the temperature of the cylinder in degrees of Fahrenheit, and T′ is the equivalent time at a standard temperature of 60°. The thermometer was noted at the beginning and ending of every set of vibrations, and was always placed in the box with the cylinder.

At all Captain Fitz-Roy's stations the apparatus was placed for observation on a stand, which raised it from two to three feet above the ground, thereby rendering the cylinder somewhat less liable to be disturbed by local influences: it was not furnished with a means of examining the strict horizontality of the cylinder, that improvement having been introduced into M. Hansteen's apparatus at a later period. On this point Captain Fitz-Roy remarks: "A small leaden tripod was used as a stand, whose upper surface was adjusted by a small spirit-level – or roughly by the trough of an artificial horizon, filled with mercury. Upon the leaden stand the box containing the needle was adjusted by its foot-screws, so that the suspending fibre of silk hung centrally in the wooden tube, the needle's centre being over that of the graduated circle, and the needle itself near, but not touching, the bottom of the box. The needle was not always strictly parallel to the bottom of the box, nor strictly horizontal, because I would not move the brass stirrup in which it was suspended, but its deviation from strict horizontality never exceeded two degrees, and was seldom nearly so much."

The time of completing every tenth vibration was recorded. The time of performing 300 vibrations is deduced from a mean generally of seven partial results: i. e. from the 0th to the 300th; 10th to the 310th; and so on, to the 60th and 360th vibrations; the commencing vibration being always at an arc of 20°. In a very few instances the number of vibrations observed, after the commencing arc of 20°, was less than 360; in such cases the first vibration after the arc had become 20° has still been taken as the commencing one, though previous ones may have been recorded; it being kept strictly in view, to obtain the relative time of vibration in arcs as nearly the same as possible, and not exceeding 20° as the initial. The arc was reduced to 10° generally about the 100th vibration; and one thousandth of the time of vibration has been taken throughout the series as the correction to infinitely small arcs.

The object of noting the time of every tenth vibration is to check errors in the counting, which will sometimes occur in the course of the 360, particularly with the very short and quick-moving needles of M. Hansteen's very portable apparatus, and at stations of low dip, where the horizontal force is greatest, and the needle consequently moves most quickly. Several such mistakes evidently occurred. When the time of completing every supposed tenth vibration is observed with tolerable exactness, and the duration of each pair of vibrations decidedly exceeds any irregularity of probable occurrence, apart from miscounting the number of vibrations, such mistakes can be discovered with ease, and rectified with certainty. This has been done in every case where no doubt could exist of a mistake of the kind having occurred; such as when all the intervals are of nearly equal duration, with one or two exceptions, which differ as much as three or four seconds from the general body. There are two stations, however, Callao and Keeling Islands, where the rectification is not so clear, or the true result so obvious. At Callao there are three series of horizontal intensities, each of forty observed intervals, which should be of ten vibrations each. Several of these intervals are between 17,5 and 18,5 seconds, and several others between 20,5 and 21,5 seconds. These can hardly represent an equal number of vibrations, because the difference between them is greater than can easily be supposed due to any uncertainty in seizing the particular beat of the chronometer at which the vibration was completed; it is, moreover, about the time that would be occupied by two vibrations more or less. The question then arises, do the longer intervals represent 12, and the shorter 10 vibrations, or do the longer represent 10, and the shorter 8? In the former supposition the intensity at Callao would be about 1.01 (Paris = 1.348): in the latter about 0.75. The difference shews how great an error would be risked by either assumption. If we take a mean of all the intervals as they stand, the amount of error risked would be certainly lessened; but we should assuredly not have the true time of three hundred vibrations, except on one supposition: namely, that the irregularities in question are not errors in estimating the number of vibrations, but that each interval really represented an equal number, and that some unusual and accidental cause occasioned the needle to differ so greatly in successive intervals. But this supposition would imply a disturbing cause vitiating the series as a measure of the magnetic intensity at the station. I have not ventured, therefore, to draw any conclusion from these observations, farther than to notice, as above, the limits within which, in either of the two first suppositions, the intensity would fall.

A nearly similar reasoning applies to the observations at Keeling Islands; of three series, one is decidedly so irregular, that no inference could be drawn from it; in the two other series the irregularities are neither so frequent, nor so large: my general impression (in the uncertainty created by the irregularity of the first series), is, that the majority of the intervals are of twelve vibrations, and not of ten: if of twelve, the intensity would be about 1,21; if of ten, about 0,85.

The inconvenience of the rapid motion of the needle, occasioned, at one part of the voyage, the practice to be discontinued of observing every tenth vibration, and every twentieth was substituted. This no doubt relieved the perplexity in which the observer occasionally found himself, in having to observe, and record, and be prepared again to observe, at every twenty seconds or less, and so far the change enabled him to observe better. But still, the disadvantage remains, in so quick moving a needle, that if a mistake of two vibrations is made, the difference of time occasioned is not of so marked and decided a character as to be at all times at once distinguished. It is of much more importance that there should be no miscount of the vibrations, than that the times should be recorded correctly to the fraction of a second. It is only the earlier and later times that are finally influential; but every undetected error in the number of vibrations falls with its whole weight upon the result.

The occasional discrepancies in the results of the same, or of different, observers, or on the same, or on different, days, which are seen in the subjoined table, are not, I believe, traceable to the source I have been discussing, nor apparently to any other than an actual difference in the time of the cylinder performing its vibration. A mean has been taken as the result at each station, except at St. Helena, where the discrepancy on the 11th and 13th of July was so considerable, that it has been thought more satisfactory to collect the observations of each day into separate results.

The subjoined table comprises the result of each observation, and the general results deduced for each station. The column entitled "Time" is that of 300 vibrations; and the "Corrected Time" is the mean of these, corrected for the rate of the chronometer and the arc, and reduced to an average temperature of 60°. The dips are those observed by Captain Fitz-Roy; except at Port Famine, where, as Captain Fitz-Roy did not observe, it has been supplied from Captain King's observations; and at Coquimbo, where, for the purpose of computing the intensity, it has been supplied by estimation from the other geographic positions on this coast, at which Captain Fitz-Roy observed the dip. In the column showing the time of vibration as a dipping-needle at Plymouth corresponding to the periods of observation at the several stations, the compensations have been introduced for the variation in the intensity of the cylinder, agreeably to what has been said above on that subject. The two final columns exhibit the values of the total magnetic intensity at the different stations derived from these observations. In the first of the two columns, the values are given relatively to the force at Plymouth, considered as unity; and in the second column, relatively to the force at Plymouth, expressed by 1.375; for the purpose of exhibiting Captain Fitz-Roy's results in direct comparison with the determinations of continental observers, who have taken Paris as their basis, giving the force at Paris the arbitrary expression of 1.3482. I have taken the ratio of the force at Plymouth to that at Paris to be as 1.375 to 1.348, which I believe will prove a very near approximation; it is that which results from Captain Fitz-Roy's observations at Plymouth, in October 1836 (page 17), and mine, at Tortington, in Sussex, in June 1837 (page 10): the dip at Tortington, at the period in question being 68° 57′, and the intensity, compared with Paris, through the medium of London, 1.368.

Captain King, having hitherto made known his observations with the same cylinder in the years 1826 to 1830 only by communicating them to M. Hansteen, from whom he received the apparatus, has now given permission to Captain Fitz-Roy to publish them with his own. I have already noticed the great loss of magnetism which took place in this cylinder during Captain King's voyage, and the care with which that officer availed himself of every opportunity of ascertaining, by direct observation, the proportion of the loss sustained in separate portions of the voyage. There are twelve stations of observation on the east and west coasts of South America, besides three stations in ports of the Atlantic on the outward voyage. By the practice of repeating observations at the same station at distant intervals, the South American stations are so linked together and connected, that by adopting a method similar to that used in determining longitudes by means of chronometers, we may compute the intensity at all the South American stations referred to and dependent on the force at Rio de Janeiro; regarding Rio in the same light as a first meridian is considered in determinations of longitude. We may then make Rio the means of connecting the whole series with Europe; for which it is remarkably well suited, the intensity there having been determined, independently of Captain King, by four observers of different nations, whose results are extremely accordant.

The dip observations of Captain King were communicated, in occasional correspondence during the voyage, to M. Hansteen, who computed them by Mayer's formula, and arranged them in a table, of which a copy was given by Captain King to Captain Fitz-Roy, and is printed in the next page. At some of the stations Captain Fitz-Roy also observed the dip in the subsequent voyage, and, as will be seen, the results of the two observers sometimes differ considerably. This may have been caused, either by instrumental or other error of observation, or by actual differences of dip existing in different localities at the same station.

Dips, observed by Capt. P. P. King, 1826 to 1831.

The following Table contains Captain King's Observations of Intensity, with the times of vibration corrected for the arc and reduced to a standard temperature of 60°.

Magnetic Intensity, observed by Capt. P. P. King between 1826 and 1831.

At Rio de Janeiro, which was the first station observed at in South America, the cylinder was vibrated in August 1826, September 1827, and December 1828; in the intervals between these dates are comprised the greater part of the observations on the east side of South America. There is no direct observation at Rio subsequently to December 1828; but we are enabled to supply the time of vibration, which would have been observed had the cylinder been employed at Rio on June 1, 1830, in the following manner. We have seen that on the 15th September 1827 the time of vibration was observed at Rio; on the 18th December following it was observed at Monte Video. These observations give the intensity at Monte Video relatively to that at Rio, subject to whatever change of magnetism the cylinder may have undergone in the interval of three months. This comparison was repeated in the following year, on Captain King's return from Monte Video to Rio, the interval being nearly of the same duration, and the order of the experiment reversed, the passage being in this instance from Monte Video to Rio, it having been before from Rio to Monte Video. On the supposition of an uniform, or nearly uniform rate of change in the cylinder, the errors arising therefrom during the two passages would be of opposite kinds, and should compensate each other in a mean of the two comparisons. Calling the force at Rio unity, these comparisons give its value at Monte Video respectively as follows, namely,

On the 1st of June 1830, being then on his return from the west coast of South America, and on the eve of sailing for England, Captain King again observed the time of vibration of the cylinder at Monte Video; whence, through the preceding comparison, we obtain the time of vibration at Rio, which should belong to the same date. We have thus a fourth date at Rio, which, added to those enumerated above, will include the whole of the South American stations; and we have only to distribute in each interval the loss of magnetism which the observations shew to have taken place from one date to the next, in the manner which may appear most suitable. There is no very obvious indication that the loss was other than gradual; and by considering it uniform in each separate interval, the results are found extremely accordant at several other stations at which observations were repeated at distant intervals.

In the first of the subjoined tables are given the times of vibration at Rio at the four periods referred to; and the corresponding times as a dipping needle. In its three last columns are shewn, – the number of days comprised in each interval, – the increase in the time of vibration owing to the loss of magnetism, – and the resulting daily correction, on the supposition in each case of the loss having been uniform in the interval during which it occurred.

The second table contains the corrected times of horizontal vibration at each of the South American stations, at the dates respectively inserted, – the corresponding times as a dipping needle, – the times of vibration as a dipping needle at Rio de Janeiro at the same dates, derived from the observations in the first table, – and the resulting intensity at each station relatively to unity at Rio. Thus far the results are derived from Captain King's observations, unmixed with those of any other observer: but in order to bring Captain King's series into connexion with the general body of results of other observers, the values of his intensities are expressed in the final column in terms of the scale in common use, in which the force at Paris = 1,348, and at Rio de Janeiro 0,884; the latter being the mean of four independent determinations by the following observers, namely,

Port St. Elena is not included in this table, as no dip was observed there, and the total intensity consequently cannot be computed. The three stations, Madeira, Teneriffe, and Port Praya, at which the cylinder was vibrated in the outward voyage, are also without dips observed by Captain King. The deficiency at Port Praya has been supplied from Captain Fitz-Roy's observations and my own, both having been made at the same place at which Captain King's intensity was observed, – Captain Fitz-Roy's at a later, and mine at an earlier date. At Madeira also the dip has been supplied from my observations, which were made in the British consul's garden at Funchal, where Captain King's cylinder was vibrated. I have deducted, from my determination of the dip, 12′ for the probable change between 1822 and 1826. At Teneriffe the dip has been frequently observed; but the values assigned by different observers vary so much as to indicate a more than usual frequency of local disturbance, which might also be expected from the geological character of that island. It would be unsafe therefore to employ any dip for that station but one which was certainly obtained at the same spot at which the horizontal intensity was observed.

The dates of the observations at these three stations fall between the observations at Greenwich in March 1826, and those at Rio de Janeiro in August of the same year. Having the intensity at Greenwich = 1,372 and at Rio = 0,884, and the dip at Greenwich 69° 52′, and at Rio 14° 00, we have the time of vibration of Captain King's cylinder as a dipping needle at Rio at the respective dates as follows, namely,

It appears therefore that but a very slight change took place in the magnetism of the cylinder during the outward voyage, and we may take 536,6 as the time of vibration at Rio, corresponding to the dates of the intermediate observations; and this is done in the table for Madeira and Port Praya.

Table I.

Table II.

4. – The following Table exhibits the Magnetic Observations of Captains King and Fitz-Roy collected in one view.

General Remarks.