ON DIMORPHOUS BODIES. 211 
dimorphism and isomerism ; there exist also other observed ana- 
logies which point to that reduction in the number of received 
elementary substances which must necessarily follow the esta- 
blishment of the supposition that elementary bodies are suscep- 
tible of isomerism. 
Thus certain compounds, like cyanogen, known by the name 
of radicals, exhibit all the chemical relations to the elementary 
bodies by which simple substances belonging to the same class 
(chlorine, bromine, &c.) are distinguished ; the latter therefore 
may likewise be compound. 
Again, the chemical and physical relations of the several states 
of isomeric bodies are sometimes (cyanogen and paracyanogen) 
at least as distinct from each other as those exhibited by the 
several elementary substances comprised in almost any of the 
natural groups*. This consideration adds weight to the hypo- 
thesis that the latter are not simple. 
71. The speculations of chemists in regard to the probable 
diminution of the number of received elementary bodies have 
hitherto run only in the channel of decomposition. Nor is 
this surprising, since up to the present time the greatest ac- 
cessions to our knowledge have flowed to us through this chan- 
nel. It has been often supposed that any given elementary 
substance A, as happened with the alkalies and earths, may 
prove to be made up of two others known or unknown; and 
that in any two of them, if the constituents prove the same, 
they may be united together in different proportions. The 
idea of a possible transformation has hitherto hardly been 
thought of ; and yet the doctrine of isomerism, rich already in 
its numerous discoveries, has shown that any number of the re- 
ceived elementary bodies may be made up of the same elements 
united in the same proportion. That they are so made up is in 
no degree the less probable, that under no circumstances have 
we ever observed any two (as iodine and bromine) to be trans- 
formed into each other, since even of the isomeric groups few 
are yet known, the members of which are mutually convertible 
by methods as yet understood or at our command. 
Regarding the question under this new point of view, it will 
appear that the study of the several kinds of physical and che- 
mical properties which the same portion of matter may assume, 
and of the circumstances which influence the development of 
one or other of these kinds, if it do not ultimately solve, is not 
unlikely to throw considerable light upon this, the most inter- 
_ * Cyanogen is not more like to paracyanogen than oxygen is to sulphur; 
less so than chlorine is to iodine. See Transactions of the Royal Society of 
Edinburgh for 1836, vol. xiv. 
P2 
