W. A. Norton —Variability of the Ultimate Molecule. 191 
cules furnished by allotropism and isomerism, we may mention 
that “many well known substances exhibit differences in hard- 
ness, color, specific gravity, solubility, etc., according to the 
circumstances in which they are produced.” d 
(3.) Some substances in the nascent state exhibit chemical 
number of their constituent atoms, or in the physical and 
mechanical condition on which their chemical activities depend, 
without any difference in the number of constituent atoms. 
The former supposition involves the improbable hypothesis 
that the same atoms regarded as endued with inberent forces of 
a constant intensity, may take up two or more different relative 
m 
positions of equilibri 
ih 
variation, might be extended almost indefinitely. In fact, in 
brought about, 
f, in view of the array of evidence that has now been pre- 
sented, it be admitted that the ultimate molecule has the 
property of variability, in the sense that has been defined, under 
varying mechanical relations to other molecules, and the mole- 
cule be regarded merely as a group of kindred atoms, then we 
must conclude that these atoms are also variable, after the same 
manner as the molecule itself. For it is obvious that a group 
of kindred atoms cannot exercise an external action varying: “4 
intensity at a given distance, and increasing eee ‘i 
the amount of change experienced in its mechanical relations 
