364 TKOF. W. J. SOLLAS ON THE MODE OE [Aug. 1 895, 



velocity at the surface across the different sections being as 

 follows : — 



30 mm. from the origin - 5 mm. per diem. 



60 „ 1-0 



90 „ 1-5 



115 „ 2-0 



It will be seen by reference to fig. 3, which represents at half the 

 natural scale the longitudinal section through the middle of the 

 ' glacier,' that the downward movement near the source, or in the 

 region corresponding to the snowfield of a true glacier, has by no 



Fig. 3. — Longitudinal median section through the 'glacier' at 

 the conclusion of the experiment. 



30 m.m. 



115 m.m. 



30 m.m. 



90 m.m. 



115 m.m. 



means been confined to the upper layers, but has affected almost 

 the whole thickness of the pitch ■ a corresponding upward move- 

 ment has taken place near the barrier, so that the layer numbered 3 

 has been lifted up and carried right over it, and layers 2 and 1 are 

 following in its wake. 



There is one particular in which pitch differs markedly from 

 glacier-ice, and this is its powerful adhesion to the sides of the con- 

 taining trough. The effect of this, however, is to produce a simi- 

 larity in behaviour, since adhesion retards the movement of the 

 bounding-surfaces of the pitch in the same manner as the friction 

 of a rough and rugged channel retards the adjacent ice. The re- 

 tardation owing to adhesion has amounted to actual arrest in the 

 ease of the lowest layers, where they touch the back of the trough, 

 as is shown in fig. 4 (p. 365), which truly represents them as not 

 having been there disturbed at all ; a retardation has also naturally 

 occurred at the sides of the trough, and the total result has been to 

 give to all the layers the spoon-shaped form that has been so 

 frequently commented on in the case of the successive layers of ice 

 of a glacier. It is this adhesion which partly explains the loop- 

 like form of the curves 1 and 2 in fig. 3, which represent in section 

 the two lowest layers. Evidently there is less slipping between 

 the pitch and its containing walls than between pitch and pitch. 



The extent to which the several layers have subsided in the 

 region 30 mm. from the origin of the flow is given in millimetres 

 in the following table : — 



