230 J.P. Cooke on the Law of Definite Proportions 
headed ‘ Mean Sp. Gr. of Zine and Antimony” are given the cal- 
culated specific gravities of the same crystals on the supposition 
the two metals had undergone no expansion on uniting. 
The last column was obtained by subtracting the numbers of the 
former from those of the latter, and therefore shows the relative 
amount of expansion. On examining the table, it will be found 
Ist. Than the union of antimony and zinc is accompanied by ex- 
pansion. 2nd. That the specific gravity of the crystals varies 
slightly with the composition. 3d. That the two minimum spe- 
cific gravities correspond precisely to the composition of Sb Znz 
and Sb Zns, so that the specific gravity increases and the expan- 
sion diminishes as you depart on either side from these two cen- 
tres. Ath. That the specific gravity of Sb Zns is smaller than 
that of Sb Zne. We find then that the specific gravity deter- 
minations confirm in general the results of the analysis pointing 
out the same two centres of crystallization. 
Crystalline Form.—It has already been stated that only two 
crystalline forms can be obtained from the alloys of zine an 
timony, that of Sb Zns and that of Sb Znz. A large number of 
crystals of Sb Zns from different alloys, and therefore containing 
different proportions of zinc, were carefully measured for the pur- 
pose of ascertaining whether the angle wasat all affected by the va- 
riation of composition. Fortunately four different crystallizations 
afforded excellent crystals, the angles of which could be measured 
toa minute. ‘T'he crystals contained respectively 43:15, 4414, 
peared to be very constant for in all cases where it could be accu- 
rately measured the same value was obtained. As none of the 
crystals of Sb Zn2, containing an excess of antimony, could be 
measured with precision, no constant variation of angle could be 
bg and on the other hand it could not be proved to be inva- 
riable. 
Affinity for Oxygen.—The affinity of the crystals of Sb Zns, 
of different compositions, for oxygen, may be estimated by com- 
paring the amounts of hydrogen gas evolved in a given time on 
boiling alloys of the same composition with water. ‘The results 
