W. Gibbs on the Atomic Weights of Elements. 249 
The first column of numbers giving the atomic weights, the sec- 
ond the atomic heats or the products of the atomic weights into 
the specific heats. It is possible that thorium belongs to the 
tetratomic group, as thoria appears to be ismorphous with stannic 
acid, Sn,O,. ‘Lead is placed in this group from the analogy of 
the compounds which it forms with ethyl, methyl, &c. to stan- 
nethyl, &c.: it appears however like tin, palladium, ruthenium, 
platinum, rhodium, iridium and osmium to be also diatomic; the 
so-called protoxyds having the formulas Pb,O,, Sn,0,, & 
For similar reasons we admit the existence of 6 triatomic ele- 
ments, namely :— 
Nitrogen, 14 — Antimony, 120 611 
Phosphorus, 31 5°85 Bismuth, 208 6°41 
Arsenic, 75 6°10 Boron, 11 a 
The diatomic elements, according to the reasoning above men- 
tioned, will be 30 in number, and may be arranged in natural 
families, as follows :— 
Oxygen, 16 + Tron, 56 «638 
Sulphur, 32 6:48 Manganese, — 
Selenium, 80 = 608 Cobalt, 60 6-42 
Tellurium, 128 6°08 ickel, 58 6°28 
Chromium, 
Magnesium, 24 — Aluminum, 275 5°88 
ium, 40 ‘Zine, 
Strontium, 88 — Cadmium, 112 6:36 
’ 136 opper. 63 6°04 
Uranium, 1 
Cerium, 92 a Mercury, 00 3=- 640 
Lanthanum, 4 oe He eae FD Eo 
Didymium, 96 one Molybdenum, 96 6°91 
Yttrium, wade ‘eum Tungsten, 184 6°67 
Erbium, shes IS pa Vanadium, ? 188 — 
Terbium, sass atic nee 
Glucinum, 10 ith 
riu 118 
This classification into natural families appears to me to repre- 
Sent the present state of our knowledge, though the re of 
Several elements must be regarded as doubtful.* ‘The mona- 
tomic elements are only 10 in number; they may be arran 
In t grou 
S3 
Chlorine, 35-6 Hyd 1 — Silver, 108 6°16 
s oor ydrogen. er, : 
Bromine, 80 -674 Lithium, 7 — Gold, 107 68 
odine, 127 6:88 Sodium, 23 6°71 
Fluorine, 19 oo Potassium, 39 6°60 
soa ttete is reason to believe that vanadium belongs with boron to the nitrogen 
