AND COMPOSITION IN OEOANIC COMPOUNDS. 
275 
of their boiling-points ; but the boiling-points of the primary monamines greatly differ 
from those of the isomeric bases of the two previous classes. 
Constant relations then between the b offing-points and the formulse of the volatile 
organic compounds may be said to have been most positively established by the pre- 
ceding statements. Their existence cannot be called in question on account of their 
not always being manifested exactly in the same manner. Such uniformity would obtain 
if in all homologous series the difference in boiling-points corresponding to the difference 
of composition C 2 were equally great ; if, as a general rule, on comparing two corre- 
sponding pairs of compounds the same difference in composition were found to involve 
the same difference in boiling-point. The fact of these relations obtaining in a less 
general and simple manner renders their perception, as also their use in determining 
the formulae of chemical compounds, more difficult, but does not set them aside, any 
more than the existence of a relation between chemical composition and crystalline 
form could be denied, on account of its not always manifesting itself in the simplest 
form, or the assistance be doubted, which the study of crystalline form often renders in 
establishing the formula of a substance, because compounds of altogether different 
atomic constitution may possess the same form, or compounds of analogous constitution, 
even containing so-called isomorphous elements, are observed to crystallize in forms 
altogether different. 
Relations between boffing-point and composition have been more especially proved 
in organic compounds ; very many of them, being volatile at comparatively low tempe- 
ratures, admit of their boffing-point being accurately determined. It is, however, in the 
nature of the case that they should not be limited to the domain of organic chemistry. 
Nevertheless relations of this description have not hitherto been comprehensively proved 
to exist in inorganic bodies. Tribromide of arsenic. As Biq (boiling-point 220°), and 
trichloride of arsenic, As CI 3 (133°), exhibit nearly the same difference as oxybromide of 
phosphorus, PO 2 Brj (195°), and oxychloride of phosphorus, PO 2 CI 3 (110°), and also nearly 
the same as the following organic compounds (exhibiting a similar difference in their 
formulae), bromoform, H Bi'g (152°), and chloroform, C 2 H CI 3 (62°). In these cases the 
substitution of 3 Br for 3 Cl is attended by an elevation of the boiling-point amounting 
to from 85° to 90°= 3 x 28 to 3 x 30. But the substitution of a^Br for x Cl by no means 
involves invariably an elevation of the boiling-point of from x 28° to x 30°. Although 
earlier observations had pointed to these simple relations, and raised the hope that a 
knowledge of the differences in the boiling-points of corresponding bromine and chlo- 
rine compounds would be available for ascertaining how many equivalents of bromine in 
a substance are substituted for chlorine in another, yet the determination of the boiling- 
points of a very large number of corresponding bromine and chlorine compounds has 
unmistakeably shown that so simple a relation does not obtain, and cannot therefore 
throw any light upon the formulae of such substances. 
The recognition then of definite relations between composition and boiling-point is 
for the present chiefly limited to organic compounds. For the majority of these com- 
