332 



Mr. J. C. McComu-l. 

 JTio. 7. 



[Mar. 12. 



that it should not slip off the projection left by the gradual evapora- 

 tion of the unsheltered surface. 



The string carrying the weight was put as close as possible to the 

 stirrup without risk of touching it, and so that the central point of 

 the aperture came somewhere between the two. 



Exp. 10. The bar was taken from the same bath ice as in 

 the last experiment. It was all one crystal with the axis vertical. 

 The first attempt was a failure, owiug, I believe, to some snow 

 getting underneath the iron plate, and, by giving way gradually, 

 tilting up the plate. I had put a good deal of snow inside the 

 cigar box, with the hope of preventing evaporation. This made the 

 readings erratic and unreliable, so the next day I turned the bar 

 over to give the stirrup a smooth surface to bear upon, and star 

 fresh. The results are given in the table (p. 333). I think 

 amount of depression may generally be trusted to within O'Ol mm. 



Several interesting points are brought out in this table. When 

 the weight is changed, the alteration in the rate of depression is great 

 out of all proportion, e.p., the alteration from 0'0058 to O410 wht-u 

 the weight is changed from 0'174 to 1*47 per square cm. During the 

 course of the experiment there was a decided rise in plasticity ; 

 compare the earlier with the later rates under 1*47 per square cm. 

 at similar temperature. This is corroborated by the increase of 

 speed under 0'85 kilo. 



The only exception, viz., the decrease of speed at first under 

 41*7 kilos, was due, I believe, to the elastic strains which had bet n 

 set up in the preliminary bending. The effect of these elastic strains 

 is shown by the undoubted rise of the middle of the bar when the 

 weight was removed at the end of the experiment. 





