VARIATION. 



But ftill, that great phenomenon, the •variation of the 

 variation, i. e. the continual change of the declination in 

 one and the fame place, which the modern obfervations daily 

 confirm, is not accountable for on any of thefe foundations, 

 nor even is it confident with them. 



Dr. Hooke communicated to the Royal Society, in 1674, 

 a theory of the variation, the fubllance of which is, that the 

 magnet has its pecuhar pole dillant ten degrees from the 

 pole of the earth, about which it moves, fo as to make a re- 

 volution in 370 years ; whence the variation (he adds) hath 

 altered of late about ten or eleven minutes everj' year, and 

 will probably fo continue to do for fome time, till it begins 

 to become flower and flower, and will at length be ftationary 

 and retrograde, and in all probabihty may return. Birch's 

 Hill, of the Royal Society, vol. iii. p. 131. 



Dr. Halley, in the Philofophical Tranfaftions, N° 148, 

 invented a new theory, founded on a great number of ob- 

 fervations, many of which were made exprefsly for the 

 purpofe by order of the government ; but as they do not 

 extend to a more recent date than about 1680, and as per- 

 petual changes are going on in the variation of different 

 places, the table of refults of this learned philofopher is 

 now of little ufe ; we (hall therefore avail ourfelves of tJie 

 hiftorv of thefe changes, as publifhed in a recent work on the 

 variation of the compafs by W. Bain, mafter in the royal 

 navy ; which contains much important information on this 

 Tubjed, and is deferving of the particular attention of every 

 one engaged in the management and navigation of veflels. 



At London, in 1580, the quantity of variation was found 

 to be 11' 15' E. ; in 1662, 6° E. ; in 1634, 4° 5' E. ; and 

 in 1 6)7, the needle coincided with the true poles of the 

 world ; fo that a period of 87 years elapfed in changing 

 the 11° 15' of eallerly variation in that city to zero, or 

 until the variation began to take a wefterly direftion. In 

 1672, the variation was 2° 30' W. ; in 1723, 14° 17' W. ; 

 in 1747, l7°4o'W. ; in 1780, 22°4l' W. ; and in 1793, 

 •vi%. 136 years after the time when the variation was zero, 

 it was nearly 24-!° W. ; and it is ftill nearly the fame. 



At Paris, in 15J0, the variation was 8^ E. ; and in 1660, 

 the needle pointed to the true poles of the world ; in 1 68 1 , 

 the variation was 2° 2' W. ; in 1760, 18^ 20' W. ; in 1804, 

 22° 20' W. Hence it follows, that whilft the variation was 

 undergoing an annual change of 10' 4" during a period of 

 213 years in London, the yearly change at Paris during a 

 term of 254 years was only 7' 10". 



At Dublin, in 1657, the needle coincided with the true 

 poles of the world ; and in 1791, the variation was 27° 23' 

 W. ; exceeding the variation obferved at London by 3° or 

 4° ; and, confequently, the annual change during 134 years, 

 muft have been about 12' 10". 



While the variation was undergoing this change at Lon- 

 don, Paris, and Dubhn ; we find very nearly a correfponding 

 change at Cape Aguillas, and at the Cape of Good Hope. 



At the former of thefe places there was no variation in 

 1600, and in 1692, it amounted to 11° W. And at the 

 Cape of Good Hope, in 1 700, the variation was nearly 

 10° W. ; and in 1791, it had increafed to 24° 31' 52" W. ; 

 fo that during a period of 91 years, the annual change in 

 the variation at the latter place, muft have been about 



At St. Helena, the variation in 1600 was 8^ E. ; in 1692, 

 1° W. ; in 1776, 13° 15' W. ; and in 1794, it was found to 

 be 16? 16' W. ; confequently, the mean annual change in the 

 variation at this place, during a period of 194 years, has 

 been at the rate of 7' 52". 



At Cape Comorin, in 1620, the variation was 14° 20' W. ; 

 in 1688, 7" 30'; in 1756, 0° 15' W. ; and in t8i6, there 



was ftill no variation at this place ; therefore, during a period 

 of 137 years, the mean annual change was & 17" ; but this 

 includes a term of 60 years, in which there appears to 

 have been a very fmall change in the declination, ria. about 

 15" annually. 



The phenomena prefented by the variation at Cape Horn 

 and its vicinity, are extremely different from thofe obferved 

 at the Cape of Good Hope, Paris, and London. 



In 1683, in S. lat. 57° 27'. W. long. 57- 28', the vari- 

 ation was found to be 23° 10' E. : in 1775, in S. lat. 56° 27'. 

 W. long. 54°, the variation was 24° 23' E. : in 1786, in 

 S. lat. 53°. W. long. 70^^, the variation was 22" 47' E. : 

 and in 1795, in S. lat. 57^. W. long. 67°, the variation 

 was exaftly 23° E. Hence it follows, that during a period 

 of 1 1 2 years, the variation near Cape Horn has neither 

 increafed nor diminilhed in a perceptible manner. 



At Cambridge, in Mafl'achufetts, in 1708, the variation 

 was found to be 9° W. ; and fince that period, it has been 

 diminiftiing at the rate of ij' annually ; whilft at Jamaica, 

 Barbadoes, and Lima, the variation has undergone no change 

 during a period of 140 years. In the northern hemifphere, 

 in the parallel of Spitzbergen, Davis's Straits, Hudfon's 

 Bay, &c. the fame quantity of variation appears to have 

 exifted during the fpace of 150 years. The quantity or 

 variation from Cape Comorin ealtward, towards Nicobar 

 iflands, Java, Ceram, Amboyna, Timor, &c. is fo very fmall, 

 as feldom to induce navigators to advert to it in their calcu- 

 lations ; and is fubjeft to little or no change. It may, 

 however, be worthy of remark, that 60 miles eaft from the 

 coaft of Coromandel, about the meridian of Madras, the 

 variation changes from eaft to weft ; and the fame local 

 changes probably take place on the coafts of Chili, Peru, 

 and Mexico. In 1704, from Valparayfo to Acapulco, a dif- 

 tance of 50° of latitude, the variation was very inconfider- 

 able ; for from the former of thefe places to Lima, it never 

 exceeds 3° E. ; and from Lima to the latter place, never 

 more than 4° W. 



The greateft variation that the author, from whom the 

 preceding abftraft has been made, is acquainted v\-ith, is 

 that which has been obferved between Cape Farewell and 

 Labrador, in Hudfon's Straits and Baffin's Bay. In N. 

 lat. 52°! W. long. 52°, it was found by the author above 

 alluded to, to be 40° 10' 53" W. ; and as high as N. 

 lat. 60°, in about the fame longitude, it was found to be 

 50° or 52° ; but he expredes fome doubt whether this great 

 excefs of variation might not have been partially produced 

 by the effedls of local attraftion. 



In 1 61 6, in N. lat. 78°. W. long. Bo° (Baffin's Bay), 

 the variation was found to be 57° W. ; and in 1757, N. 

 lat. 62°. W. long. 65°, the variation was 41° W. ; and in 

 the fame year, in N. lat. 63°. W. long. 79°, the variation 

 was 43° W. : and in all thefe high latitudes, the variation 

 ftill continues nearly the fame. 



On the weft fide of America, in the fame parallel of la- 

 titude as Davis's Straits, Cape Farewell, &c. we perceive 

 the variation affuming another charafter, and feldom ex- 

 ceeding half the quantity found at the above-mentioned 

 places. 



In 1786, in N. lat. 53°. W. long 145° 21' from Paris, 

 the variarion was frotn 23° to 24° E. ; in the fame year, in 

 N. lat. 58i°. W. long. 138°, it was 25° E. ; and in Port 

 des Frangois, in N. lat. 58° 37'. W. long. 137° 30', the va- 

 riation, as afcertained by the meridian line, amounted to 

 27° E. In 1793, in N. lat. 53°. W. long. 1 29°, the variation 

 was only 20° 41' E. ; and in 1794, in N. lat. 61° 17'. W. 

 long. 149° 7', 29° 30'eafterly variation was found the greateft 

 quantity obferved by Vancouver while on that coaft. 



7 But 



