38 SECTIONAL ADDRESSES. 
build up a crystalline lattice. Crystals and aggregates of crystals are 
thus the only true solids, glasses being regarded as under-cooled liquids 
of high viscosity. Since the early beginnings of the geometry of crystals 
due to Haiiy, the study of their geometrical form has reached a remarkable 
state of perfection, all the possible lattices have been determined, and there 
is perfect agreement among crystallographers as to the classification of 
forms and the optical properties of different types of crystals. The 
X-ray method developed by Laue and by W. H. and W. L. Bragg has 
carried the matter an important stage further, by making it possible to 
determine, not only the class of a crystal, but the exact lattice possible 
to crystals belonging to that class. The connection between the chemical 
properties and the crystalline structure still remains indeterminate, 
although it must be very intimate. I shall revert to this point later. 
There are many reasons why the chemical study of solids should 
receive greater attention. In metallurgy, although metals and alloys 
may, and most frequently do, pass through a molten stage in the course 
of their manufacture, they may undergo many important changes of 
structure and constitution at temperatures far below that at which the 
last liquid portions have completely solidified, and these changes may be 
so far-reaching as to convert an alloy into one seemingly of an entirely 
different class, although the gross chemical composition has not altered. 
The petrologist, especially when dealing with igneous and metamorphic 
rocks, has to consider reactions which proceed in the midst of solids of high 
rigidity. Several industries, such as that of cement, are based on reactions 
of the same kind as those with which the petrologist has to deal. Sintering 
is not always due to the presence of small quantities of molten material 
between the solid particles, and it is now certain that union of solid masses 
under pressure may occur without actual melting. This was shown by 
Spring forty years ago, but for long, although frequently quoted, his 
results received little consideration. The most striking application of the 
principle is seen in the metallurgy of tungsten. This metal was formerly 
described as very hard and brittle, and it is not possible, by casting it and 
then annealing, to bring it to a ductile form. The method now adopted 
is to prepare it in the form of a pure powder, and then to bring it to a 
compact state by compressing, heating, and hammering while very hot, 
and finally drawing. As this process is continued, and as an originally 
thick rod becomes extended into a slender wire, the brittleness progressively 
disappears, and at last the tungsten is obtained in those beautiful filaments, 
drawn to extreme fineness, with which we are familiar in our electric light 
bulbs and wireless valves. Even several of the common metals, when 
their powders are compressed under suitable conditions at temperatures 
far below their melting points, are capable of forming compact masses 
with a mechanical strength of the same order as that of the cast metal. 
The conditions of these reactions, which have been studied by Sauerwald, 
suggest interesting questions for consideration. A somewhat similar, 
but perhaps more difficult, problem is that of the adhesion of an 
electrolytically deposited metal to its support, which is sometimes so 
perfect as to approach the breaking strength of one of the metals although 
interpenetration of crystals is not to be seen under the microscope.. 
There is another aspect of the chemistry of solids which will make an 
appeal to some who are not chemists, but amateur students of Nature. 
