of Metal Rods with rounded ends. 
261 
and then by allowing single rods of different lengths to impinge 
on the plane face of a large block of cast iron which was intended 
to represent an infinite solid. The straight line law now became 
apparent, but broke down when the rods became too long, owing 
to the reflected waves in the block coming into play. It was this 
which gave the hint to use the rods in pairs. It is, of course, 
impossible to imitate an infinite solid exactly ; but if we allow two 
equal similar bodies to impinge symmetrically on each other, the 
effect on either is (by the very symmetry of the case) the same as 
if it had struck an absolutely rigid plane ; and the case of an 
absolutely rigid plane is even simpler than that of the infinite 
(elastic) solid. The duration of the impact will, of course, depend 
only on the relative velocity of the two bodies, and in the actual 
experiments one rod was allowed to hang freely at rest, the other 
impinging upon it. The impact was regulated by first withdraw- 
ing the second rod to a distance which could be measured by 
means of a telescope mounted on a horizontal travelling screw, 
and then allowing it to fall freely. After some experiments a 
mechanical release was devised which effected this most satis- 
factorily, and which, when once adjusted, enabled the same with- 
drawal to be repeated rapidly without reference to the telescope. 
A great many experiments had to be made before the 
apparatus was got to work satisfactorily, and the necessary correc- 
tions determined. These preliminary experiments were all made 
with steel rods, and with a withdrawal of 8" (corresponding to a 
velocity of impact of about 20" per sec.). When they were 
completed it was found that all the rods had become overstrained 
at the point of impact, so that the curvature there was slightly 
diminished. It was easy to detect this by observing the reflexion 
of a straight line in the polished surfaces of the ends, when a slight 
kink at once became apparent at the centre. The result of this 
flattening of the ends of the rods was to produce a diminution in 
the end-effect ; but, as the different pairs of rods had received 
greatly different treatment, this diminution was not the same in 
all cases. It was therefore necessary to re-turn and polish the 
ends, and to repeat all the experiments with a smaller velocity of 
impact. The velocity of 20" per sec., hitherto used, gives rise to 
a maximum total pressure, during impact, of about 300 lbs., which, 
on Hertz’s theory, would produce at the centre of the circle of 
contact, a pressure of about 170 tons per sq. in. This pressure 
varies as the cube root of the velocity, so that it cannot be much 
reduced without a very great reduction in the withdrawal (to 
which the velocity is proportional). It was found, however, that 
if the withdrawal were much less than 2", experimental errors 
began to increase, probably on account of the elasticity of the 
suspension. The height of fall corresponding to this withdrawal is 
