228 The American Geologist. October. 1902. 
It has been shown by Prof. Johnson * that brittle mater- 
ials fail under a compressive load, by shearing on definite 
angles, and that the resistance to move along these angles 
is composed of two parts. "First, the strength of the mater- 
ial to resist shearing, and second, the frictional resistance to 
motion along the plane." Moreover, he say^, "The relation 
of crushing strength to relative dimensions of specimens is 
a very important matter. Hitherto nearly all crushing test 
specimens of brittle materials have a cubical form. So long 
as the theoretical angle of rupture was thought to be 45°, 
this was proper, but since this theoretical angle approaches 
60°, it is evident that the height of the specimen should be at 
least one and one-half times the least lateral dimension, in 
order to allow of failure on a normal angle.'" 
Professor Johnson gives a mathematical demonstration 
01 this important conclusion, which is also verified by the 
results of experimental tests. It is shown in the case of some 
limestone blocks which were tested, that when the cubical 
form was used, the results were 9 per cent greater than would 
have been true if the specimens had been chosen of the proper 
hight. He says, "The unit strength of the material is no 
function of the si::e of the spcciineiis. but oii!\\a functiou of 
its form," and that "Crushing test specimens should be true 
prisi>is in form." The work of professor Johnson did not in- 
clude experiments with ice specimens, yet it is noticeable that 
the characteristic appearance of the ruptured materials, as re- 
presented by half-tone cuts, appears to coincide with the de- 
scril^ed appearance of the ruptured ice blocks. 
It is to be regretted that no photographic representations 
of the broken ice specimens are given by any of the experi- 
menters. Since, however the more compact specimens 
of ice "broke suddenly with report," it is probable that the ice 
behaved strictly as a brittle substance at the lower temper- 
atures. This being true, the apparent strength of the culiical 
blocks as tabulated may have lieen, on the whole, nearly one- 
tenth too high. This suggestion is made here, not so nuich to 
criticise the work referred to. upon the crushing strength of 
ice. as to indicate to future ex])erimenters a ])Ossiblc elimination 
of error. 
■""Brittle materials, under compressive stress." J. B. Johnson. Di- 
gest Of Physical Tests. 1S96-97. 
