72 Scientific Intelligence. 
basis for all comparisons of molecular refractive powers, but the 
results are nearly as satisfactory when deduced from the sae: 
observed with the red light of the hydrogen isan. 
values 7, were thus obtained. 
If now we compare the molecular dept powers calculated 
from the atomic composition (R,) with those deduced from direct 
observations on the densities and indices of refraction of the com- 
Mall 
pounds—M L)—we shall find that in the case of a large num- 
of organic products the generalization of Landolt is very 
aaa confirmed, thus 
TABLE III. 
(" * ) R, Diff 
Propylalcohol C;H,0 28°6 28°4 +0°2 
Propylaldehyde C;H,0 26-0 25°8 +0.2 
Propylethyl ether C;H,20 43°8 43°6 +0.2 
Propyla CsHi002 44-0 44-0 00 
Propylchloride C;H,Cl 34/1 33°9 +02 
Isobutyric acid C,Hs0. 36°3 36-4 —O1 
ex: oHi4 48°6 48°2 +04 
Tri-ethylamine CeHisN 55-3 55:3 0-0 
If then the molecular refractive power is simply the sum of the 
atomic refractive powers it must be the same for all isomeric 
howe es and therefore independent of their molecular structure, and 
so cage ae thought. In 1870, however, it was shown by Glad- 
stone (London Chem. Soc. Jour., viii, 147) that there were many 
y th i 
ously, when ni eis atoms are united b e than a “eee bond, 
feature in common. Hence all isomeric bodies have the same 
molecular fate. power, ane the differences of molecular 
structure do not extend to the relation we have named ; and the 
discrepancies observed fe Gladstone arose from the circumstance 
that in the molecules of the bodies he chiefly studied two or more 
of the ~~ n atoms were linked by double or treble bonds. 
aring the molecular refractive powers of bodies con- 
taining guably or cable linked multivalent atoms, after the 
