The researches which I beg, in the following pages, to submit to the Royal Society, embody the results obtained in the further development of an observation which I made a considerable number of years ago, and which, since that time, I had to defend against the objections of others, both by experimental inquiries of my own, and by the collection and discussion of facts elicited in the investigations of other observers. As far back as 1841* I pointed out that in analogous compounds the same difference of composition frequently involves the same difference in boiling-points. The assertion of the existence of this law-like relation between the chemical composition of substances and one of their most important physical properties, when first enunciated, met rather with the opposition than with the assent of chemists. In Germany especially it was contested by Schröder in his memoir “On the Molecular Volume of Chemical Compounds.” These objections led me to collect additional evidence in favour of my views, and to show more particularly that in very extensive series of compounds (alcohols C
n
H
n+2
O
2
; acids C
n
H
n
O
4
; compound ethers C
n
H
n
O
4
, &c.) an elementary difference
x
C
2
H
2
is attended by a difference of
x
X 19°C. in the boiling-points, and how this fact is intimately connected with other regularities exhibited by the boiling-points of organic compounds. Almost at the same period Schröder § convinced himself that the relation I had pointed out obtains in most cases. He collected himself a considerable number of illustrations of the regularities I had traced, and showed that the relation in question is rendered more especially conspicuous if the compounds be expressed by formulæ representing equal vapour-volumes of the several substances. Some of the views, however, which were peculiar to Schröder have not gained the approbation of chemists. This physicist was inclined to consider the boiling-point of a substance as the most essential criterion of its proximate constituents, as the most trustworthy indicator of its molecular constitution. His views were chiefly based upon the assumption that the elementary difference C
2
H
2
, when occurring in alcohols C
n
H
n+2
O
2
, involved a difference of boiling-points other than that occasioned by the same elementary difference obtaining in acids C
n
H
n
O
4
and that the isomeric compound ethers differed from one another in their boiling-points. An extensive series of boiling-point determinations* which I made of these isomeric ethers, proved that the latter assumption is not founded on facts. The exertions made by Schröder, Gerhardt, Löwig and others, in the hope of recognizing the influence of the constituent elements on the boiling-point of a compound, have also essentially remained without result.
A new model for predicting boiling points of alkanes
