52 SECTIONAL ADDRESSES. 
highly suggestive in regard to the study of solid solutions by chemical 
means. In all such work it is important to remember that the size of 
the crystals, the possibility of the material having been cold worked 
previously to testing, and other physical and mechanical factors must 
be taken into account. The properties of single crystals in this connection 
are unknown, and the preparation of single crystals of solid solutions is 
much more difficult than that of pure metals, so that further work will 
be required before any definite opinion can be given as to the validity of 
Tammann’s conclusions. 
As a consequence of the early studies of crystals by means of X-rays, 
some metallurgists were at first disposed to accept the conclusion that the 
chemical molecule ceased to exist in the solid state. This generalisation, 
which was instinctively felt by chemists to be improbable, was premature. 
Crystals of elements are clearly so constructed that all atoms are similarly 
related to one another, and there is then no group intermediate between 
the atom and the whole crystal, whilst crystalline salts are best regarded 
as built up of ions, every chlorine atom in rock salt being equally related 
‘to six sodium atoms and so forth, but these conditions do not exhaust 
the possibilities. Organic compounds undoubtedly retain the chemical 
molecule, or some simple multiple of it, in the solid state, and the same 
is true of the very interesting class of compounds which metals form 
with one another. These are of a non-polar character, and hence have 
long puzzled chemists on account of their utter disregard of valency. 
Such a compound as NaHg, melts at 360°, or more than 260° above its 
less fusible component, and is largely undissociated in the molten con- 
dition. It evidently represents a very stable union of the sodium and 
mercury atoms, and it has many analogues. The intermetallic compounds 
have several features of interest. Their space lattice arrangement has 
been studied in a number of instances, but the correlation of their chemical 
properties with their atomic and crystalline structure still remains to be 
undertaken. 
If our knowledge of the chemical properties of the interior of a crystal 
be very incomplete, what are we to say of its surface? Of this we know 
still less. ven in a crystal of a pure metal there must be some difference 
in the structure at the immediate surface, on account of the unsymmetrical 
forces between the atoms in the outermost layer and its neighbours. For 
so far as the radius of sensible atomic forces extends, therefore, there must 
be a condition different from that which prevails at a depth below the 
surface. One consequence is that the surface has residual affinity, which 
shows itself in the ease with which foreign atoms or ions will attach them- 
selves to it. That the forces acting are chemical is shown by the great 
effect on the extent of adsorption of the chemical character of the solid 
and of the adsorbed substance. Films, often one atom thick, attach 
themselves to the solid, and are only removed with the greatest difficulty. 
Their presence makes the investigation of the properties of a surface 
difficult, as the surface actually examined may be in reality quite different 
from that which is assumed to be present. In photochemical experiments 
with mercury it is usual to prepare a completely fresh surface of the liquid 
metal by causing it to flow continuously in a fountain, but this device 
cannot be applied to solids. Only rarely can experiments be made with 
perfectly defined solid surfaces. Films of metal prepared by sublimation 
