ME. EOBEET MALLET ON VOLCANIC ENEEGY. 
183 
103. Great numbers of experiments have been made by engineers and architects upon 
the resistance to crushing forces of various rocks employed in building See., some of 
which, such as those of Gautiiey and Rondelet, are of great exactness ; but unfortunately 
these are nearly all inapplicable to our purpose : — (1) because those experiments being- 
made with a view to structural purposes, the recorded crushing-pressure is very commonly 
taken as that at which the specimen of rock first begins to crack or give way ; (2) because 
no reliable record is found of the height through which the surface producing the 
crushing-pressure has descended between its level of first application and that of the 
mass of powder produced, so that we have an unreliable and always too small value for 
W, and none at all for h ; (3) the experiments recorded have been generally made on 
very small specimens (cubes of 1 centimetre or of 1 inch on the edge), and of rocks the 
lithological characters of which are imperfectly handed down. 
104. It was therefore necessary to institute a completely new and independent series of 
experiments, and upon as large a scale as that for which competent apparatus could be 
procured. 
The rocks to be experimented upon being reduced by the marble mason or lapidary’s 
tools to exact cubes, or parallelopipeds as nearly cubes as possible, it is obvious that 
experiments for determining the work of crushing may be conducted either, as in the 
writer’s preliminary ones, by the free descent of a weight just enough to crush, or by 
the steady increase of a load until the crushing in each case has occurred. 
The latter is greatly to be preferred, not only as avoiding some conceivable sources of 
error in crushing by impact, but for the convenience afforded by the increase of a steady 
pressure up to crushing when operating upon a large variety of different rocks. 
105. Through the obliging kindness and zeal for science of John Ramsbottom, Esq., 
Engineer, and by permission of the Directors of the London and North-Western Railway, 
the writer had fitted for his purpose, and placed with a staff of men at his disposal, a 
magnificent machine constructed from Mr. Ramsbottom’s design for the locomotive works 
at Crewe ; and not only to those gentlemen, but to W. M. Moorsom, C.E., of the same- 
works, are his best thanks due for the very efficient assistance in every way afforded him. 
106. The testing-machine at Crewe thus employed consists of a large wrought iron 
balanced lever, as seen in Plate IX., so constructed that it can be applied to com- 
pression, tension, or torsion, the load being produced by the flow of water into an iron 
cylindric vessel suspended to the long arm of the lever, the weight of which at known 
temperatures is registered by an index at each instant. This simple form of testing- 
machine possesses great advantages in accuracy and certainty, as to the load actually 
visited upon the object, over any of the complex machines in which the production of 
the load or its registration is conducted through a series of connected levers. 
By shifting the fulcrum the lever could be altered in power at pleasure from 10:1 
to 20 : 1, and its own weight could be balanced so as, if desired, to form no part of the 
load visited on the specimen. The strength of the parts is sufficient to admit of a 
crushing-strain of SO or 90 tons with safety. The largest size was chosen for the cubes 
of stones to be crushed that this limit w T ould with safety allow. A cast-iron cage or frame, 
