180 F. W. Clarke on the Atomic Volume of Compounds. 
gle observations have been made, it is clear that hats variation 
from equality may be ascribed to errors in experim e 
extreme difference in atomic volume is from a mininium ‘of 43°6 
to a maximum of 50°7. It is not rare for an equal divergence 
to occur between different determinations for a single substance. 
Again, although chlorids, bromids, and iodids have unequal 
atomic volumes, those of similar chlorates, oe and 
iodates are equal,—at least as far as we have any dat 
Shu ae sp. er : ‘289, Bodeker. NaBrO,, 3:339, an NalO,, 
4277, mers. At. vols., 465, 45-2, and 4 
KCIOs,, sp. st, 3 305, Buignet. KBrOs, 3-271, and E10, 3°979, 
Kremers. At vols., 52°7, 51°1, and 53:8. 
Most similar compounds of silver and sodium have equal 
atomic volumes. ‘This holds true of the chlorids, bromids, and 
iodids (as will be shown in another connection hereafter), and 
of the sulphates, chlorates, nitrates, and probably carbonates. 
The sulphids and oxyds are exceptions. 
Some compounds of As, Sb, and Bi, have equal values. 
As,03, sp. gr., 3°695, Gathoutt: 3°884, Filhol. Sp,Os, 5°11, Terreil ; 
5'78, Bo ullay. BhOy 8-079, P. & J.; 8450, Leroyer & 
Dumas, At. vols, 51 053° 6; 50 5—5T- 5 Be ‘and 65:4— 
57°9. 
In the carbon group compare certain oxyds. 
Si0,, 2-663, Deville. TiO, (artificial anatase) 3°700, Hautefeu- 
ille. SnO, 6720, Daubrée. At. vols., 22-5, 22 2 and 22°3. 
Would space permit, I might go on multiplying examples to 
an almost indefinite extent; but my object at present is merely 
to illustrate certain principles. There are other regularities yet 
to be noticed. The first of these is comparatively unimportant, 
and I will merely state it as it is, without adducing any evi- 
dence. All sulphates have atomic volumes which are lower than 
the sum of the metallic oxyds andthe SO,. That is, the value 
for MnSO, is less than the sum of those of MnO and SO, To 
this rule I have found no exceptions. Whether a similar rule 
holds good for chromates, molybdates, and tungstates, I am not 
certain; the materials at my command being too limited to set- 
tle the question. 
nother regularity is more remarkable. It is ws 
that certain oxyds and sulphids have atomic v 
are lower than those of the metals contained in pe Now, in 
a number of cases, I find that. those atomic volumes equal ha 
the sum of those of the metal and the other element, “provided 
that S=15°6, and O receives its highest value, 10-4. I give oxy- 
gen this value, for this reason. ~ Many sulphids have atomic 
volumes, as I have already stated, equal to the sum of those of 
