456 Dr. W. W. J. Nicol on 
was found that separation of Na, SO, 10aq. was attended by 
expansion, Na, 8, O3 5aq. by contraction ; and it was this con- 
tradictory behaviour of the two salts that caused me to abandon 
the experiments for the time. 
In more recent experiments | employed the apparatus in 
the figure, which is in many respects more accurate and more 
easily worked with than a bottle; at the same time I saw the 
futility of determining the density after crystallization: the 
whole secret was to be found in the solution before solidifi- 
cation. iixpansion or contraction was a mere accident 
peculiar to the salt employed. 
Table I. contains the results of all my experiments calcu- 
lated for anhydrous salt-molecules per hundred water-molecules 
with the corresponding apparent molecular volumes (=) 
nr 
of each salt-molecule. 
TABLE I. 
n( Nas S, Os) 100 H,0. 
ree n. | Density. Mol. vol. — 
[e} 
20 7665x | 1:38896 | 2167-84 47-99 
Xn 9°216 1:43835 2263'80 50°33 
e 10:057 146231 pa alot | 51:46 
3 le oyal 1-49392 2357-70 5201 
‘. 11-679 Fo0136 ||. 242803. | 53°77 
a 12-132 | 61250383 a 247 1b. e| 55:35? 
“ 15-149 bo ltd9006 i 2637-37 3| 55:29 
. 20°000t | 1:673835 | 2964-20 58°21 
20 5 | 1:03466 1808-32 | 16:64 
* 1:0 106744 1819-31 19:31 
- 1:847* 111783 1845-77 24°71 
i 3193 oy AOD AS 1890°18 28°25 
# 4-442 1:25855 1931-40 29°58 
re 4-729 1:26698 1969°35 31:87 
5 6°244 134014 200471 32°79 
* Saturated at 20°. + Fused crystals. 
In the case of sodium thiosulphate, the first solution deter- 
mined was that saturated for the hydrated salt at 20° C.. and 
after that solutions of various strengths up to that consisting 
of the salt fused in its water of crystallization. With sodium 
sulphate the solutions ranged in strength from the half-mole- 
cule up to one containing more than six molecules of salt per 
