C. W.Andrews — On Kerater}:)etuni Galvani, Euxley. 81 



versely the ice on the summits would have a tendency to thin. If 

 the erection have parapets of small height, when the ice-stream 

 thickening rises higher than the parapets, parts will ultimately 

 break away. This answers to the Aletsch glacier flowing past the 

 opening of the Marjelen See valley. But if the parapets were very 

 high, 1 think that ultimately the hollows must fill up to at least 

 the levels of the summits, and so produce an umlulating but generally 

 descending slope. Such, indeed, we see in actual glaciers, if there 

 be rock-basins hidden in their beds. 



I have not sought the figures for the viscosity and breaking-strain 

 of ice. Blocks in a frost can lie about without perceptible change; 

 the roof of the Esquimaux's snow-hut does not sink down on his 

 head during the night ; the walls of the Montreal winter-castle 

 remained as long as they were wanted. Perhaps ice does not begin 

 a viscous flow till the force reaches a sufficient amount. A stick of 

 sealing-wax may lie on its side for months without noticeable 

 alteration, but " hang a weight to a bar of it, and it will yield " 

 (Lord Kelvin, "Popular Addresses," vol. ii. p. 344:). 



Practically we see that the Swiss and Greenland glaciers do 

 move as viscous bodies. They follow paths of least resistance, and 

 these paths are much the same as water would follow. I have 

 been asked — What should happen when a glacier from a side valley 

 debouches on to the top of a larger one in a main valley ? Much 

 the same as what happens when a tributary flows into a main river, 

 as may be seen where the glacier from under the Sparren-horn 

 debouches on to the Aletsch. Water is viscous and ice is viscous, 

 though in very different degrees. " Water does not run uphill," 

 we say, but in a pipe it does. Ice can move uphill if it must, but 

 we do not find it flowing uphill if there be an equally eligible down- 

 bill path available. This is the argument of a letter of Professor 

 Bonney's in Nature (22nd February, 1894), which met with little 

 reply. The subject tempts me to a further pursuit, but 1 will not 

 transgress the limits which I indicated in beginning. 



YI. — Note on a Specimen of Keraterpetum Galvani, Huxley, 

 FROM Staffordshire. 



By C. W. Andrews, B.A., B.Sc, F.G.S. 

 of the British Museum (Natural History) . 



IN a collection of Coal-measure Vertebrata made by Mr. J. Ward, 

 of Longton, and recently acquired by the British Museum, 

 there is an imperfect, crushed skeleton of a small Lab3annthodont 

 from the Ash-coal Shale of Longton Hall Colliery, Staffordshire. 

 This specimen was noticed by Miall in 1874 in his British Associa- 

 tion Report on the Classification of the Labyrinthodonts, where it 

 is referred to JJrocordijliis, some of the characters given in his 

 diagnosis of that genus being taken from it. Some measurements 

 of the skull were added. The same specimen was afterwards 

 figured in Mr. Ward's paper " On the Geological Features of the 

 North Stafi'ordshire Coal-field" (Trans. N. Staffs. Instit. Mining 



DECADE IV. — VOL. II. — NO. II. 6 



