12 AN INVESTIGATION INTO THE 
a diagrammatic view of the apparatus, which consists of a pair of brass tubes, 
B, B f , provided with set screws, A, A', for attachment to the specimen, and 
connected together by a flexible steel plate, F, forming the fulcrum. The 
ends of the tubes near the fulcrum plate are pressed apart by an adjustable 
spring S, to insure a uniform pressure on the screw points gripping the 
specimen. On the opposite end of the tubes is a spring finger, D, of ebony, 
pressing against a double knife-edge, K, seated in a shallow V notch cut in the 
end of the other arm. The knife-edge carries an adjustable mirror, M, so 
that if any change in the diameter of the specimen occurs the two tubes 
move relatively to one another in a horizontal plane and thereby cause 
the knife-edge mirror to rotate; the rotation of this latter is observed and 
measured by a telescope and scale placed at a suitable distance. 
For convenience in adjustment there is a screw, L, for tiltifrg the apparatus 
about the axis of the gripping screws, and the tubes B, B' are trussed to 
prevent vibration. This instrument was calibrated by aid of a Whitworth 
'measuring machine and the scale adjusted so that one division corre- 
sponded to one-millionth of an inch. 
APPLICATION OF THE METHOD OF SIMPLE COMPRESSION TO THE 
DETERMINATION OF THE CUBIC COMPRESSIBILITY OF METALS. 
The behavior of such metals as wrought iron and steel over a wide range 
of stress shows' that these metals may be considered as almost perfectly 
elastic. The results of the theory of elastic bodies may therefore be applied 
in their cases with great confidence. 
As a typical example of the behavior of such materials we may consider 
the deportment of a specimen of wrought iron when subjected to a cycle 
of compression stresses, commencing at 1,000 pounds and rising to 9,000 
pounds, afterwards returning to the original load. 
The readings obtained for the longitudinal and lateral strains will show 
in such a case that equal increments or decrements of load produce 
strains which are very exactly proportional thereto. This is clearly shown 
in a plot of these readings, where the ordinates represent the total load and 
the abscissae represent strains. In both cases the relation of stress to strain 
is represented by a straight line returning upon itself. Traces which vary 
but little from the ideal straight line are given by black Belgian marble, as 
will be seen on page 25. 
Such results afford an arbitrary standard by which can be judged the degree 
of approximation to perfect elasticity exhibited by other metals and by rocks 
under similar conditions. 
If we now calculate the value of the modulus E for simple compression, 
since this is the relation of the compression stress p to the strain e, we have 
p=Ee 
