6 SEVENTH REPORT—1837. 
fore M. de Humboldt’s departure ; but as the needle was left in 
Mexico, those observations could not be made on the return to 
Europe, by which its magnetic invariability might have been 
assured. ‘lhe circumstances are greatly to be regretted, what- 
ever they may have been, which deprived a suite of observa- 
tions so extensive, and on which so much care and labour had 
been bestowed, of a final confirmation, which can hardly be 
supplied in an equally satisfactory degree by any less direct 
_ evidence. Fortunately, indirect means are not altogether want- 
ing in this case, and we may infer from them that up to the 
beginning of 1800 M. de Humboldt’s needle had undergone no 
change; and that if subsequently to that date it lost magnet- 
ism, the alteration was not considerable. The observations in 
Paris were made in 1798. Between August 1799 and February 
1800, M. de Humboldt made thirteen determinations of the 
intensity on the Spanish main, between the latitudes of 10° and 
11°, and the longitudes of 292} and 2963. The mean of these 
is an intensity of 1:196. In 1822 the value of the intensity at 
Trinidad, in lat. 10° 39! and long. 2984, was determined, by 
observations made by myself (to be discussed hereafter), to be 
1-204. The result of this comparison is extremely satisfac- 
tory; and being derived, on M. de Humboldt’s side, from obser- 
vations with one needle at several stations, and on mine from 
several needles at one station, a fair conclusion may be drawn, 
that in the beginning of 1800 his needle retained its magnet- 
ism unimpaired. In January, 1801, M. de Humboldt’s needle 
gave for the intensity at Havannah 1°359; mine, in 1822, 1-499. 
In this comparison the agreement is less perfect; there is a 
greater difference than is usual between the results of different 
observers at the same station ; and it is such as would be occa- 
sioned by a loss of magnetism in M.de Humboldt’s needle, but 
not to an amount that would impair in a material degree the 
value of his important series. Against any precise inference, 
however, to be drawn from these comparisons, there is, Ist, 
the difference of the dates_at which the respective intensities 
were determined ; 2nd, a small difference in longitude of the 
localities of the first comparison ; and 3rd, those circumstances 
of a local and instrumental nature which must affect every 
such comparison. 
In the account which M. de Humboldt has given of his ob- 
servations there is no mention made of corrections having been 
applied for the arcs of vibration or for the temperature of the 
needle; but in such an extensive series, corrections on these 
accounts are of minor importance. 
The number of Jand-stations at which the intensity was ob- 
a ee oe 
