﻿M. A. Heim on Glaciers. 499 



any length of time in a thickly liquid state. Dried by heat, it 

 becomes full of cracks and mealy. Such a gypsum paste the 

 potters and plasterers call " killed." 



Killed gypsum therefore consists of rod-like and needle-shaped 

 crystals only loosely adherent by the water in the interspaces ; 

 it is a mass of wet powder. 



By the addition of more or less water, any degree of thick 

 liquidity can be exhibited. The experiments on the motion of 

 killed gypsum were conducted in the following manner. Out 

 of potter's clay a glacier-valley, or several joined together were 

 prepared, with windings and more and less steep, narrower and 

 wider parts, about 4 feet long; and then the gypsum paste was 

 poured in at the upper part, and this put into flowing motion 

 as slow as was desired by knocking at the base or by raising the 

 upper end. With paste slightly liquid the motion was a simple 

 flowing ; the more thickly liquid it was made, or the smoother 

 the valley- sides, the more was the motion a compound of sliding 

 and flowing ; and it could easily be employed so thick that the 

 motion exhibited only a sliding of the whole mass along the 

 valley-sides. Corresponding to the conditions in glaciers, a 

 mean was preserved, which effected a motion compounded of 

 sliding and flowing. The fissures produced by non-uniform 

 motion are often as fine as a hair. If the water with which the 

 gypsum is killed be previously coloured darker (for example, 

 blue), these figures will all be still more evident, being filled 

 with the coloured water, while the white gypsum will be only 

 tinged. The, whole impression given by these gypsum streams 

 is perfectly that of a glacier. My honoured teachers, Professor 

 Escher von der Linth and Professor Moussou, have viewed some 

 of the experiments. The following are some of the results. 



The Middle Wall of Compound Glaciers. — If nearly equal 

 streams of gypsum be caused to flow through two valleys into a 

 third formed by their junction, immediately at the confluence a 

 depression is produced solely by the mechanical force of the 

 motion, but a very little further on a decided central wall-like 

 elevation which soon attains a constant height, and only after 

 a longer course gradually disappears again. Each of the two 

 streams was semicircular in transverse section, about 5 centims. 

 broad, and 2 centims. deep ; the united stream was 8 centims. 

 broad. The resulting central longitudinal wall had at its most 

 perfectly formed part a breadth of about 13 millims. and a 

 height of 3 millims. The phenomenon is often very striking, 

 but does not always occur in this manner ; it is very much influ- 

 enced by the angle of junction of the streams, the form of their 

 transverse section, &c. 



This immediately suggested the conjecture that a part of the 



