TRANSACTIONS OF SECTION B 989 
seca what the arrangement at each step is, and thus to trace a connection 
tween the chemical constitution and the crystalline form, &c. 
Thus in iron pyrites, FeS,, the hemihedral form with parallel faces which is 
characteristic of this mineral, and which is from it called the pyritohedron, can be 
traced on from the chemical molecule FeS, through a macromolecule which 
includes six of these as sub-molecules, and which is connected with the other 
similar macromolecules in a regular way. 
So, again, the general form and tetrahedral hemihedralism which are found 
in boracite can be traced from its formula, 
3Mg0, 4B,0,, 
through macromolecules each of which contains four of these chemical molecules. 
And, again, the two kinds of hemihedralism, the right-handed and the left- 
handed, which present themselves in different specimens of quartz, can be traced 
from the chemical molecule SiO, through a macromolecule which contains six of 
these connected in one or other of two possible ways, and thence on to the two 
crystalline forms of the perfect crystal that present themselves. 
There is therefore reason to believe that in each of these crystals the macro- 
molecule or crystalline molecule consists of several chemical molecules, and must 
be carefully distinguished from the latter, containing in iron pyrites six, in boracite 
four, and in quartz six, chemical molecules. 
8. On the Dilatancy of Media composed of Rigid Particles in Contact. 
By Professor OspornE Reynops, F’.R.S.—See Section A, p. 895. 
9. On the Evidence deducible from the Study of Salts. 
By Spencer U. PICKERING. 
In this paper the author dealt with the evidence as to the molecular weights of 
salts, derived from a study of the composition (1) of hydrated salts; (2) of basic 
salts; (3) of double salts, He also criticised the evidence deducible from experi- 
ments on hydration, dehydration, and the vapour tension of hydrated salts, and 
finally examined the conclusions drawn from the calorimetric investigations of such 
compounds. The conclusions arrived at by the author are, that, although ina few 
isolated cases the molecular weight obtained would appear to be greater than the 
analytical results necessitate, still in a vast majority of cases there are no grounds 
for multiplying these weights, and, indeed, there is a considerable mass of evidence 
in favour of adhering to the simplest possible formule. Such a conclusion may, at 
first sight, appear opposed to conclusions drawn from other sources. On the one 
hand the author considers it undeniable that if we succeed in determining the 
number of replaceable portions of the elements in any compound, we determine 
ex hypothesi the number of atoms in the molecule—that is the molecular weight— 
and although the data‘at our disposal at present are of the most meagre description, 
they nevertheless seem to point incontestibly to the simplicity of these molecules. 
On the other hand, considerations based on the crystalline form and other physical 
properties of bodies force on us the conclusion that liquid and solid molecules are 
in all probability of a very complicated nature, certaimly more complicated than 
gaseous molecules. Both these conclusions the author considers to be reconcilable 
with one another, and contends that because the smallest particle of a substance 
which enters into a chemical reaction may be simple, there can be no reason why 
many of these particles may not agglomerate and act in unison as regards certain 
physical forces. That this agglomerate does not act as a unit towards chemical 
forces would simply imply that the force which unites the individuals constituting 
it is not chemical force, or is chemical force of sucha weak nature that, in presence 
of the strong chemical agents we make use of, it is inappreciable. The molecule 
of achemist is not necessarily identical with the molecule of a physicist. 
