Molecular Constants. 331 
On this point, without making a study of the specific 
gravity of alloys of sodium of different strengths, I have satis- 
fied myself that, as long as the amalgam is liquid, it is lighter 
than mercury. This is easily shown by introducing mercury 
into one limb and the various liquid amalgams of sodium into 
the other limb of a long U-tube: whereupon the pure mercury 
always prevails in weight. Now when a solid amalgam of 
sodium is brought into contact with mercury, heat may be 
either set free or absorbed. Chemists will understand me if I 
remind them that a pounds of water mixed with b pounds of 
chloride of calcium will give a body which will set free or 
absorb heat according as a is greater or less than x. 
§ 27. Accordingly I made a pound or two of a sodium amal- 
gam of such a strength as to be solid at the atmospheric tem- 
perature. This was beaten up in an iron mortar as it cooled. 
Putting some of this into a porcelain crucible, plunging it into 
water containing a few drops of hydrochloric acid, and collect- 
ing the hydrogen, it was found that after a day or two, if the 
amalgam was occasionally stirred, all evolution of hydrogen 
ceased; the volumes, reduced to dry hydrogen at 0° C. and 
760 millim., were < -.^q > cub. centim. The mercury, after 
drying, was found to weigh < 1 3.7841' "^ s & Yes * ne P er_ 
centage of the amalgam which I shall call Am amalgam: — 
Hg 98-2 97-97 98*08 
Na J/8_ 2-03 1-92 
100-0 100-00 100-00 (mean) 
The ideal amalgam would perhaps be one of such a compo- 
sition that heat would neither be set free nor absorbed on 
further mixing with mercury. But such an ideal condition 
could only be ideal in its beginning, and, I think, disturb- 
ances due to this cause are insensible in comparison with other 
sources of error. The above amalgam when stirred with mer- 
cury may reduce its temperature as much as 5° C. 
I am informed that sodium may contain a large quantity of 
hydrogen. I am not called on to discuss the experiments (not 
my own) upon which this rests; but I think that any consider- 
able quantity would be expelled on amalgamation. Perhaps 
the glow or blush to be described immediately and in § 28 is 
due to the escape of residual hydrogen at the released tension- 
surface of the mercury. 
The first experiment in regard to the diffusion of sodium out 
of this amalgam into mercury was of course a qualitative one. 
A U-tube (fig. 5, PI. V.) was made of glass tube of | inch in- 
ternal diameter — the one limb, A, being about 3 inches and the 
