342 Professor Forbes' Account of his recent 



of the earlier cosmogonists, based upon somewhat vague ana- 

 logies with better understood phenomena, as when the analogy 

 of magnetic attractions seemed to offer a parallel in the me- 

 chanism of the heavens in the theory of Gilbert, and that of 

 fluid currents gave rise to the Cartesian vortices. The New- 

 tonian theory was based upon its coincidence with the empiri- 

 cal laws of planetary motion. We have as yet no empirical laws 

 of glacier motion, consequently no proper mechanical theory 

 can as yet be adequately tested. I endeavoured to point out 

 in my lectures how a mechanical theory might be deduced 

 from observation, and how these observations might be prac- 

 tically made. I believe that I have also obtained for the first 

 time, the numbers on whose importance I insisted. I am not 

 aware that any one had hitherto proposed to determine the 

 diurnal velocity of a given point of a glacier with reference to 

 three co-ordinates. The analogy with the empirical laws of 

 astronomy is both striking and just ; an exact acquaintance 

 with the path described by any molecule of a glacier, will 

 almost as certainly lead to a knowledge of the cause of its 

 motion, as the theory of gravitation sprung from the three laws 

 of Kepler. We have to deal, indeed, wdth an effect more com- 

 plex and varied ; but the results contained in my last letter, 

 already shew how much of numerical precision may be attained. 

 I have already determined the diurnal motion of 10 points of 

 the Mer de Glace with a probable error, not exceeding, I think, 

 a quarter of an inch in any case ; and when these obsen^ations 

 shall have been pursued, as I expect to do, until the end of 

 September, there will be a tolerable basis for sound specula- 

 tion. 



In particular, you will recollect that I pointed out last winter 

 two experiments for distinguishing between the prevailing 

 theories of De Saussure and De Charpentier, those of gravita- 

 tion and of dilatation. One was the exact measurement of a 

 space along the ice to be measured after a certain time, in 

 order to ascertain whether any expansion had occurred. The 

 other was the determination of the linear velocity of the glacier 

 at any point, which, on the theory of Saussure, ought (if the 

 glacier be of nearly uniform section) to be uniform throughout; 

 on the theory of Charpentier it ought to increase from 



