C. 8. Peirce on Chemical Theory of Interpenetration. 79 
6. The Thermal Equivalents of Isomorphous Crystals. 
7. Kopp’s Law of Boiling points. How is this explained? 
8. Prout’s Law as modified by Dumas. 
The only atomic weights which have been determined with 
sufficient accuracy to test the law, beside those of Stas, are the 
following :— 
Carbon 6°01 Berzelius; 6°00 Dumas and Stas; 6:00 Erdmann 
and Marchand; 6-06 Liebig and Redtenbacher; 6:03 Strecker. 
C is not more than 6-004. 
Lithium, Dieh| 7-026 (prob. error +°006); Troost 7:01, Mallet 
(S=16-03, Na=23-05, Mg=12-0125) 7-027. Mean 7:02. 
Calcium 20-°002 (C=6-004) Erdmann and Marchand. 
ith less accuracy we have 
fron, Svanberg and Norlin (after rejecting two discordant ex- 
periments according to Peirce’s criterion) 28°048; Berzelius, 
28-024; Erdmann and Marchand, 28-012; Maumené, 28°000. 
Mean 28-017. 
Combining the first three atomic weights with those deter- 
mined by Stas, we have:— 
Experiment. Law. Difference. | Dif.--E «p. 
K 39°154 39°25 | —*096 00 
Na 23°05 23 +05 500 
Ag | 107-94 | 108 ~06 | rhe 
Pb 103°45 103°5 —"05 FOOD 
Cl 35°46 355 | —04 gto 
14:04 14. | +04 wis 
s 16-03 16 +-03 sehey 
1005 1 +005 Pain 
Li 7-02 7 +02 zoo 
Ca 20-002 | 20 +002 | ve dun 
6004 | 6 ‘| +004| shes 
Kis an unexplained anomaly, but the probability of only one 
difference out of thirteen being greater than *%* is 0000087, 
While the effect of the residual influence which carries K out of — 
this limit is only ;,;, of the atomic weight. Omitting X, the 
Sum of the above differences is +°001; the probability of this 
bile ne Small is 085; hence, upon this consideration, the pro 
ability of the law is ‘782. 
ie probability is, therefore, still in favor of the law. The 
whe and they may be made still smaller by making the unit by 
18 law presents another example of the connection between 
lcal equivalents and integral numbers, and must he nd 
| ; it is 
