W. A. Norton—Force of Effective Molecular Action. 849 
inci greater than 4°934, the molecules are in general 
brought into a certain condition answerin g to the liquid state, 
and liquefaction ensues. When, by a still further rise of tem- 
perature k equals 4°934, the liquid has reached the boiling 
point in vacuo. The curve of effective molecular action is now 
that shown in fig. 8 for £=4-93, and the distance between 
the molecular envelopes, is 2°84 r 
All special curves, answering to particular solids or liquids, ie 
state. a) 0 of these is that the curves differ little from a 
right line at the neutral i a (fig. 1). This corresponds, 
graphically, to the well known law of molecular displacement, 
noe the effective resistance developed is, for small displace- 
ents, porportional to the displacement. The ratio of the 
effective force 2s, to the displacement a2, may be taken as the 
measure of the coefficient of elasticity, in considering the 
is a conspicuous result of the fiiecwisitien of the equation—that 
as the tensile stress increases, the coefficient of elasticity . 
measured by the ratio just stated, should diminish slowly a 
rst and then more rapidly. Baporiaient has established tna 
in general the coefficient of elasticity of a material varies after 
is manner. But to make the test more decisive, I have made 
a series of detailed comparisons of the theoretical with ex- 
perimental results. It appears that for all values of & rang- 
ing from 7-576 to 20 (which, as will hereafter appear, ma 
be regarded as including all the more tenacious solids) the 
law of variation of the molecular ratio, zp (fig. 1); from the 
point @ to m (i. e. from zero of stress to the point of rupture) is 
sensibly the same. Thus at the point m, answering to rupture, 
this ratio becomes reduced to 0°303 of its value at the neutral 
point, a, when k=20; to 0°301 when 4=12-41; and to 0°316, 
when k=7576; and. the correspondence is equally close at 
points intermediate between a and m, ave computed the 
= 
comparative values of the ratio —s, for eighteen supposed val- 
ues of the displacement, a2, atid ‘aac this scale of com- 
puted values,—which, as we have j just seen, should answer to any 
