14 



LIEUT.-COLONEL SABINE ON TERRESTRIAL MAGNETISM. 



the Sulphur proceeded to the islands of Socorro and Clarion, to Martins Island one 

 of the Marquesas, and lastly to Bow Island, with which the stations on the west coast 

 of America and its adjacent islands may be considered to have terminated. 



To have made this series of magnetic determinations thoroughly complete, the 

 needles should have been taken back to Panama, and their times of vibration should 

 have been re-examined there at the close of the operations ; but this proceeding did 

 not consist with other duties. We are, therefore, without that direct evidence of the 

 steady magnetism of the needles, subsequently to the observations at Panama in 

 March 1839, which might have been furnished thereby; but where so many needles 

 are employed, evidence of scarcely inferior weight may be obtained by their inter- 

 comparison ; especially at stations where the opportunities of observation are favour- 

 able, and the probable error of the result with each needle is further diminished by 

 its being derived from repetitions on different days. The observations at Mazatlan 

 and San Bias, on the return of Nos. 7- 8. and 9. from England, furnish one good oc- 

 casion of this nature ; and we may take as a second the observations at Martins 

 Island, being the last station at which they were repeated on dilferent days. 



If we divide the squares of the times of vibration of the several needles at Panama 

 by the squares of their times of vibration at Mazatlan, we obtain quotients, which, if 

 the needles were unchanged relatively to each other in the interim, should be iden- 

 tical ; or as nearly so as the ordinary errors of observation permit, including therein 

 the diurnal and irregular variations of the magnetic force itself. The times of vibra- 

 tion at San Bias and Panama, similarly treated, supply a similar comparison, in which, 

 however, the quotients will differ in absolute value from the preceding ones, inas- 

 much as the horizontal intensity of the earth's magnetism is not precisely the same 

 at Mazatlan and San Bias* ; but the degree of accordance with each other of the se- 

 cond series of quotients will furnish, as in the former case, the required evidence ; 

 which is of greater weight in the instance of San Bias than in that of Mazatlan, be- 

 cause the times of vibration at San Bias were derived from observations on two dif- 

 ferent days, and at Mazatlan from those of a single day only. 



* The quotients are, in fact, in the two cases, the respective values of the horizontal intensity at Mazatlan 

 and San Bias relatively to the force at Panama taken as unity. 



t These are arbitrary weights, assigned according to the number of days employed in each comparison. 



