OBJECTIONS TO THEORY OF VISCOSITY. 213 



The extreme fragility of the ice is difficult to harmonize with the idea 

 of viscosity. It was noticeable that whenever the ice passed over an 

 undulation of even moderate dimensions it was abundantly crevassed. 

 The movement of the ice in most such instances was obviously exceed- 

 ingly slow, so that the tension brought to bear upon the surfiice by the 

 small curve was relatively slight and came into action with exceeding 

 slowness. If the property of stretching were possessed in any but the 

 very slightest degree it would seem that crevassing would be avoided. 

 This objection, which was long since forcefully urged by Tyndall, be- 

 comes intensified when applied to the broad, slowly moving ice-sheets 

 of the far north. 



Lieutenant Peary called my attention to a glacier on the south side of 

 Inglefield gulf which breaks entirely in two in passing a steep descent, 

 and reunites below and moves on. Similar phenomena are well known, 

 but they become more emphatic in this northern region. 



I saw no indication that bowlders descend through the ice as heav}' 

 substances descend through viscous bodies. As already remarked, the 

 lamina? on api)roaching a l)owlder usually divide and a part curves under 

 and a i)art curves over it. Nowhere was seen any indication that the 

 bowlders had carried the lamina) down, as the superior specific gravity 

 of the bowlder might be expected to do in a viscous Ijody. 



Everywhere the aspect of the ice was that of rigidity rather than viscous 

 fluency. The rigidity, to be sure, did not prevent contortions and fold- 

 ings of the laminations, such as take place in cr3^stalline rocks, but fault- 

 ing and vein structures also occur, and there seems no more occasion to 

 assume viscosity in the one case than in the other. Even if a certain 

 measure of viscosity be admitted, it does not follow that viscosity was an 

 essential agency of motion. 



There is a theoretic objection to the assumption of viscous flowage in 

 the very fact of crystallization itself. The property of viscous flowage 

 rests upon the relative indifference of a particle as to its special point of 

 adhesion to its neighbor particle. The property of crystallization rests 

 upon the strongest preferences respecting such relationship. Particles of 

 water in their fluent condition lie against and cohere to each other in- 

 diff"erently. When they take on a crystalline form they arrange them- 

 selves in specific relationships by the exercise of a force of the highest 

 order. In the presence of this very forceful disposition of the particles 

 to retain fixed relationships to each other, it would seem little less than 

 a contradiction of terms to attribute to them viscous flowage. The crys- 

 talline body may readily be made to change its form by the removal of 

 particles from one portion by melting and their attachment at other 



