SEPTEMBER 18, 1902] 
the chemical dissolution of ethane, but the double mark of 
similarity between it and other methyl compounds in their chemi- 
cal interactions. Wecannot say that only one part of the ethane 
is methyl, or hydrogen, or carbon, but that part of its nature, 
of its constitution, is its behaviour as a methyl compound, or, 
again, as an ethyl compound ; or, more comprehensively but 
less specifically, part of its constitution is its behaviour as a 
hydrocarbon, as a hydrogen and as a carbon compound. But 
these are different aspects of it, different relations of it, not 
differing parts of the one homogeneous substance. 
With the laudable object of combating the prevalent notion 
that matter is something which is the basis or essence of a body, 
something acting as the medium of the manifestation of its forms 
of energy, a distinguished and most lucid writer on chemistry 
has, adequately perhaps for that object, represented a body as a 
compound of the various forms of energy subsisting together and 
cohering in certain proportions within the volume of the body. 
But this presentation of a subject as a cohesion or association of 
forms of energy is on the same footing as the presentation of 
ethane as consisting of two methyls bonded together, or two 
portions of carbon with six of hydrogen. It is compounding 
what cannot be had apart, what cannot be even conceived of as 
separate, so far as bodies are concerned. The analysis of bodies 
into manifestations of different properties are only mental opera- 
tions. A moving body, a hot body, a green body, an explosive 
body, becomes by legitimate abstraction a phenomenon of motion, 
of heat, of colour or of light, or a chemical phenomenon as our 
needs require ; but the body is there all the while, and its un- 
divided and continuous existence is indispensable to the phe- 
nomenon. The body can be hotter or colder, but not that only 
—not that without other differences ; red-hot iron is throughout 
a very different thing from cold iron, and ice differs widely from 
steam in most of its properties. A substance is no more com- 
posed of its properties or energies than it is composed of its so- 
called elements. It manifests its presence in a thousand and 
one ways more or less distinguishable ; its properties are so to 
manifest itself. But no divisibility of itself while it remains it- 
self can be thought of, no differentiation can be suggested, no 
nucleus with its superinduced properties can be traced. 
It ought, therefore, to be possible to express all the particulars 
of chemical constitution without making any assumption as to 
substances having parts or structure. Of chemical constitution it- 
self, I doubt whether there is to be found a definition which is not 
couched in language having reference to the minute mechanical 
structure of substances, notwithstanding the fact that all know- 
ledge of their chemical constitution has come to us through 
observation of the properties of the substances themselves, and 
more particularly their relations in cases of double decomposi- 
tion. Bearing in mind that all terms are relative, I think the 
chemical constitution of a substance may be defined as the re- 
semblances shown by it in its chemical changes to other sub- 
stances, often better known than it and taken as types, these 
resemblances being indicated and described usually by means of 
special nomenclature and notation. As this nomenclature and 
notation have been developed out of those designed to express 
chemical composition, it is well to point out that the notion of 
chemical constitution is independent of that of the latter, though 
clothed to some extent in its language and symbols. 
The notions of radical and atom are so intimately related as 
to be often used indifferently, the one for the other. The 
radical ethylene is always an atom of ethylene, the radical 
nitrogen always an atom of nitrogen. Radical and atom are, in 
fact, the qualitative and quantitative aspects of the same thing. 
They are thus exactly parallel with substance and molecule. 
We can think of unquantified substance, and perhaps of 
unquantified radical, but in chemistry we never really want such 
conceptions; one of the many definitions of science is the 
quantification of phenomena, and in every chemical phenomenon 
the substances concerned are quantified as molecules. The 
quantification of radicals expressed by the atom is fundamentally 
the same in principle as that of substances, namely, that of 
chemical equality in interaction ; but it may be better to say 
that it is dependent upon the quantification of substances as 
molecules. 
In the interaction of double decomposition each substance by 
contact and union with the other develops and manifests a dual 
character by becoming distributed as the two new substances, 
with the consequence that each of these has certain properties 
the same as those of the one, and certain others the same as 
those of the second interacting substance. What is common in 
NO. 1716, VOL. 66] 
NATURE 
50. 
this way to one of the interacting and one of the resulting 
substances is a radical of these substances, of which there are 
evidently four in every double decomposition. These radicals 
of a single interaction are defined as whatever two parts of the 
powers of a substance to yield the simple substances of its 
chemical composition are, in certain interactions, continued 
separately from each other in the two new substances. But the 
pair of radicals developed in the various double decompositions 
of a substance being by no means always the same, one of the 
radicals of one pair must include in its composition part or all of 
one of those of another pair. Acetic acid has for one pair of 
radicals methyl and carboxyl, and for another pair acetyl and 
hydroxyl. Of these, carboxyl includes hydroxyl and acetyl 
includes methyl. Again, acetic acid yields the hydrogen and 
acetate radicals in one interaction, and hydroxyl and acetyl in 
another, so that in these cases the acetate radical includes 
acetyl and the hydroxyl includes the radical hydrogen. Now, 
what is common to carboxyl and acetyl and what is common 
to the acetate radical and hydroxyl are also treated as radicals, 
the one being known as carbonyl and the other as the radical 
oxygen. These are examples of what may be distinguished from 
the others as the polyvalent radicals. They are radicals of 
radicals, and therefore also radicals of substances. They may 
be defined as the common part of two or more other radicals. A 
single definition of all radicals can be given, but it is not 
instructive. A radical is any single power or any interdependent 
association of the powers of a substance to produce simple 
substances which continue in any product or series of successive 
products of its chemical change. 
Before I leave the subject of the radical I wish to repeat that 
it is only when it is interacting that a substance shows a dual 
character or division, as it were, into parts or radicals, and that 
the duality it then shows is determined as much by the nature of 
the other substance as by his own. A substance is neither 
actually nor conceptually the sum of its radicals. The very fact 
of the difference of these in different interactions should be 
proof of this ; though it only leads to its being taken to be at 
least the sum of its ultimate or simple radicals. If, however, it 
is not the sum of its proximate radicals, it is hard to see how it 
can be imagined to be that of the ultimate ones. In relation to 
its radicals, a substance must be held to present itself as any 
one of these for the purpose of investigation, and at the stand- 
point from which it is considered. It is then to the mind that 
particular radical, though also something else ; just as snow is 
white and cold, yet also something else, for the moment uncon- 
sidered. Nor can the two products of an interaction be looked 
upon as themselves the sum in properties of the interacting 
substances. To a limited extent, andimperfectly, we can attach 
to a given radical certain of the properties common to its com- 
pounds ; but it needs no greater insight than we have already 
to recognise that a substance cannot be what it is in one way, 
without being in that way greatly affected by what it is in 
another. This is now a recognised but not sufficiently con- 
sidered point, and I therefore welcome those publications of 
Prof. Vorlaender, of Halle (who now honours this Section 
with his presence), in which he has been vigorously calling 
attention to the extent to which the properties of a substance, 
acid, basic, stable, and what not, depend as much as, if not 
more, upon the interrelations of the radicals than upon the 
radicals themselves. 
One other thing I have to say about the radical, which is as to 
the spelling of the word. I plead for a return to the ending of 
the word radical with ‘‘al,’’ now interdicted in the /ownal of 
the Chemical Society. It seems appropriate to call the powers 
of a substance to behave chemically as it does, the roots or 
radical parts of its chemical nature, but it does not seem 
appropriate to call them radicles or rootlets. Americans and 
all other nationalities but our own use the original spelling. 
I have put off too long, perhaps, all reference to the proper- 
ties of very dilute aqueous solutions of salts, but I wished first 
to discuss the nature of the radical. The osmotic pressure and 
other dependent points which are particular in the behaviour of 
such solutions are in full accordance with the assumption that an 
electrolyte by dissolution in much water becomes a pair or 
a binary system of two interdiffused quasi-substances called 
‘‘ions.” These ions must differ from isolated substances in 
bearing equal and opposite quantities of electricity; in being 
each unknown apart from its fellow; and in having a composi- 
tion not to be found in actual substances, though identical 
possibly with that which a radical would have were it a 
