71 
of Edinburgh , Session 1869-70. 
that the substances are under similar physical conditions. We 
cannot, therefore, expect the same uniformity in the results ; hut 
although, strictly speaking, we may entertain grave doubt on the 
real value of the results, yet, in some cases, we cannot help recog- 
nising some curious analogies, especially on comparing similar 
classes of compounds. It is not the object of this note either to 
criticise or discuss the labours and speculations of others, no 
originality being claimed in the subject matter itself, all that is 
original being merely the addition of a few new analogies. 
The first important discovery in the subject of atomic volumes 
was made by Schroter. He observed that the equivalent volume 
of oxygen, obtained by subtracting the volume of metal in the free 
istate from the volume of the oxide, gave, approximately, the same 
value of 5-2 in the oxides of copper, zinc, cadmium, lead, mercury, 
iron, cobalt, and titanium. In other words, the oxygen occupied the 
same volume in each combination. Other classes of oxides gave a 
volume of twice, or half the above number. In order to arrive at 
the volume of the oxygen, Schroter started with the premises that 
the metal in the combined state occupied the same volume as the 
uncombined metal. Granting, for the present, that oxygen has 
a definite volume in combination in the oxides, it is clear that the 
volume obtained by difference will vary with the volume of the 
combined metal. The same method applied to the oxides of the 
less dense metals would give a negative volume to the oxygen ; 
and in these cases we must admit condensation to have taken 
place in the metal itself. We may have three cases, therefore, 
according as the volume of the combined metal differs from 
that of the uncombined. If it remains the same in combina- 
tion, we obtain the real volume; if it condenses, the volume 
is a minimum ; if it expands, a maximum. Seeing that the 
oxygen in the dense metals has the volume 5-2, we may regard 
the greater and smaller volume obtained from some oxides as the 
result of condensation or expansion of the metal. Supposing the 
above volume (5*2) to exist generally in the oxides, we would 
have a condensation in the less dense metals in combination, 
approaching very nearly, in the case of potassium, sodium, and 
aluminium, to one-third, and in calcium, magnesium, and strontium 
to nearly one-half, of the volume in the free state. Thus far, 
