xv.] FORBES' GLACIER DISCOFERIES. 509 



knew very well that such analogies had no claim to found 

 a theory. I knew that the onus of the proof lay with 

 the theorist, (I), To show that (contrary to the then 

 received opinion) the centre of a glacier moves fastest ; 

 and (2), to prove from direct experiment that the matter 

 of a glacier is plastic on a great scale, a fact which seems 

 so repugnant to first impressions as lately to have been 

 urged in a most respectable quarter, 1 as rendering the 

 doctrine of semi-fluid motion untenable. No one had, a 

 right to maintain the theory of fluid motion as more than 

 a conjecture, until at least these preliminary obstacles 

 \\L re removed by direct observations/ 



' These observations have been made, and the result is 

 the viscous or plastic theory of glaciers, as depending 

 essentially on the three following classes of facts, all 

 of which were ascertained for the first time by ob- 

 servations in 1842, of which the proofs are contained in 

 this work/ 



' 1. That the different portions of any transverse section 

 of a glacier move with varying velocities, and fastest in 

 the centre/ 



' 2. That those circumstances which increase thejluidity 

 of a glacier namely, heat and wet invariably accelerate 

 its motion/ 



' 3. That the structural surfaces occasioned by fissures 

 which have traversed the interior of the ice, are also the 

 surfaces of maximum tension in a semi-solid or plastic 

 mass, lying in an inclined channel/ 



Or again (Travels, p. 365), where he finally sums up 

 his ideas : 



' My theory of Glacier Motion then is this : A 

 GLACIER is AN IMPERFECT FLUID, OR A viscous BODY, 



WHICH IS URGED DOWN SLOPES OF A CERTAIN INCLINA- 

 TION i;v THE MUTUAL PRESSURE OF ITS PARTS/ 



-e, again, may be accepted even now as a nearly 

 complete statement of the physical circumstances of tin- 

 question, in so far at least as tin -\ 'Inscribe accumtt ly the 

 observed phenomena, and account for many of them in a 

 1 < BibliotMqut UnivtrtclU, January 1843.' 



