490 MR. A. E. TUTTON ON THE THERMAL DEFORMATION OF THE CRYSTALLISED 
The amount of the expansion is relatively large, compared with the expansions 
of metals, being four times that of platinum and one and a halftimes as great as the 
large expansion of aluminium. 
The differences between the amounts of linear expansion along the three axial 
directions of any one salt are small, the difference between the maximum and 
minimum being about twelve per cent, of the total expansion in the case of each salt. 
But the differences between the values for the same direction of the three salts are 
much smaller ; in the case of the direction of the axis b, the difference is only one per 
cent., and in that of the axis a where the greatest divergence is shown, it is only six 
per cent, of the total amount of expansion. 
The increment of the coefficient of expansion, per degree of temperature, is about 
twice as large for. the direction of the axis c of each salt as for the other two axial 
directions, for ivhich the increment is nearly identical. 
It is interesting to point out, in connection with the last fact, that it agrees in a 
remarkable manner with the observation previously recorded (‘Journ. Chem. Soc.,’ 
Trans., 1894, p. 715, §14), that the change of optical refractive power brought about 
by rise of temperature is considerably greater for the direction of the axis c, than 
for the directions of the a and b axes, along which the amount of change is approxi¬ 
mately the same. 
The fact that the differences of expansion exhibited by the three salts are so small, 
compared with the differences in the amounts of expansion in the three axial 
directions, would render it probable that if any considerable changes were introduced 
in the relations of the values for these three directions l^y the replacement of one metal 
by another, particularly if such changes were not simply proportional to the atomic 
weight of the metal but expressed by a higher function of the atomic weight, such 
change would suffice to negative the possibility of a direct progression of the linear 
coefficients of expansion for each axial direction of the three salts, corresponding 
to the atomic weights of the metals present. That simple proportionality to the 
atomic weight was not to be expected directionally, has been indicated by the whole 
of the morphological and physical work on both sulphates and selenates. In the case 
of the refractive indices, however, the directional changes were small in comparison 
with the differences exhibited by different salts ; bence in the case of these, as of 
other, optical constants, the interesting progression according to the atomic weight 
of the metal was not interfered with. But although such perturbations of a direc¬ 
tional character are able, in the case of the thermal constants, to obliterate such a 
progression of the linear coefficient of expansion, they would mutually compensate 
each other when the total solid change was considered. Hence it must be apparent 
that if the influence of atomic weight were indeed a progressive one it would only 
be clearly revealed in the case of the coefficients of cubical expansion. These 
latter constants have been shown to exhibit such a progression in the clearest 
possible manner. 
In accordance with the a Dove considerations, a progression of the linear coefficients 
