'242 
MOVING FORCE. 
Cases of difS- penetration into A. If we were to suppose A and B to yield 
doc\mies*^of explanation of the phenomena as when 
moving force. A only is supposed to be penetrated, will strictly apply 3 only 
the diagram would be a little more complicated 
Let us now suppose the first part of the operation in the 
collision of A against B, to be the same as already described 
in the case of a soft body, and supposing them to be in the situ- 
ation as represented at No. 2, let us observe what must follow. 
When A has arrived opposite to F, as represented at No. 3, 
c will have returned to its original place with respect to A, 
and B will have arrived opposite toG (FG being =c EF), A 
will be at rest, and B will have acquired the full velocity v. 
Now, it is obvious, that if A had not moved on from its posi- 
tion No. 2, c would, in this last part of the operation, have 
acted upon B only till it arrived opposite to L (FL being 
= f EF), and its final velocity would have been only ^ f v 
But A having moved on to its place No, 3, c will have acted on 
B till it has arrived opposite to Gj and the force which has 
been lost by A in its passage through the space = KF, as well 
as the force of c through a space = HK, has been comniuni • 
cated to B. In other words, the force which, in the first part 
of the operation, had been expended in producing the change 
of figure, has, in the last part of the operation, been repro- 
duced by the expansion of the figure to its original state, and 
has, together with the remaining force of A, been communi- 
cated to B, If this explanation be applied to the change of 
motion produced in C and D fig. 7> as referred to at page 185, 
it must be obvious, I think, that when C is brought to rest, the 
force which it has lost, and the force of the spring, have both 
been communicated to D, 
1 In the collision of unequal masses, the distribution of the 
force is rather more complicated. Let M (fig, 15) be im- 
moveable, and filled with a dense elastic fluid, so that N, 
moving with the velocity v, and meeting with an uniform 
resistance, would be brought to rest by driving the cylinder C 
up to O. Then, if we suppose M, = 2N, to be in free space, ' 
and if we divide OP, = OR, into nine equal parts, and make 
OS 
