B.—CHEMISTRY. Bl 
; 
‘ A new field of investigation has been opened up by Tammann in his 
attempts to determine the arrangement of the atoms in solid solutions by 
purely chemical means, by studying the action of chemical reagents on 
the solid. It is a familiar fact that the ‘ parting’ of silver and gold in 
assaying, which consists in dissolving out the silver from the alloy by 
means of nitric or sulphuric acid, is only possible when the silver forms 
more than 60 per cent. of the alloy. When gold is present in excess of 
this proportion, only a little silver is removed from the surface, and the 
action then comes to a standstill, the acid being unable to penetrate to 
the interior. Assuming the alloy to be completely crystalline, the atoms 
of silver and gold will occupy the points of the space lattice, and as the 
two metals have face-centred lattices of only slightly differing dimensions, 
the amount of distortion will be small. There are, however, different 
ways of arranging the two kinds of atoms. They may be distributed at 
random, or they may be so regularly arranged as to form two inter- 
penetrating cubic lattices. The two forms of distribution may be dis- 
tinguished by means of the X-rays, but Tammann has also drawn 
conclusions on the point from the action of various reagents on the alloys. | 
He finds that each reagent which attacks silver ceases to act on the alloys 
when the proportion of gold atoms in solution exceeds a certain limit, 
which is not the same for different reagents, but which he states to be 
always capable of being expressed as 1/8, 2/8, 3/8, &c., of the total number 
of atoms. The limits so found are not consistent with the distribution 
according to the laws of probability, but they may be accounted for by a 
regular distribution on the assumption that a certain number of inactive 
atoms is necessary to protect each atom of silver. The varying action of 
different reagents depends on the number of silver atoms which react 
with each molecule of the reagent. Thus, nitric acid attacks single atoms 
of silver, solutions of sulphides need two silver atoms, whilst osmium 
tetrachloride requires four. On the basis of these results, an ingenious 
theory of the action of reagents on solid solutions has been constructed, 
_ and although the accuracy of the experimentally determined limits is not 
__ high, and there are several exceptions to the rules, an interesting case has 
_ been made out. Similar limits are found in the precipitating action of 
alloys on salts of less electropositive metals, and in the electrolytic potential 
of alloys. 
Considerations of this kind point to the possibility of a new form of 
isomerism among solids, due to the differing arrangements of the same 
atoms on a space lattice. It is claimed that such instances have been 
found. Alloys prepared by the simultaneous electrolytic decomposition 
of two metals have different chemical properties and different potentials 
from the alloys of the same composition prepared by fusion, the former 
indicating a random distribution of the two species of atoms in the lattice, 
and the latter an ordered one. Annealing of the first causes the structure 
to pass over into the stable, regular arrangement. Other properties may 
also be used as a test. Crystals of sodium chloride containing a small 
aon (0-064 mol.) of silver chloride, when prepared from solution, 
_ teadily become purple in light, whilst crystals of the same composition 
‘prepared by fusion are permanent for months. It is, however, quite 
_ possible that the former are not closely packed, and most of the facts cited 
by Tammann are capable of other explanations, but the hypothesis is 
E2 
