408 
NATURE 
[ Sept. 18, 1873 
~ those which were then on record have shown us that certain 
atoms area little heavier or lighter than was then believed, and 
the work of perfecting our observations is constantly going on 
with the aid of better instruments and methods of operation. 
But, apart from these special corrections, a more sweeping 
change has taken place, not in consequence of more accurate 
experiments interpreted in the usual way, butin cons<quence of 
a more comprehensive view of the best experimental results 
which had been obtained, and a more consistent interpretation 
of them. Thus the atomic weight of carbon had been fixed at 6 
by Dumas’s admirable experiments ; and it was quite conceivable 
that a still more perfect determination might slightly increase or 
diminish this number. But those who introduced the more 
sweeping change asserted in substance that two of these sup- 
posed atoms, whatever may be the precise weight of each, always 
are together and neyer separate from one another ; and they 
accordingly applied the term atom to that indivisible mass of 
carbon weighing twice as much as a carbon atom had been sup- 
posed to weigh. So also with regard to other elements, it has 
been shown that many atoms are really twice as heavy as had 
been supposed, according to the original interpretation of the 
best experiments. This change was brought about by what I 
may be permitted to call the operation of stock-taking. Dalton 
first took stock of our quantitative facts in a bus’ness-like manner ; 
but the amount and yariety of our chemical stock increased so 
enormously after his time, that the second stock-taking absorbed 
the labours of several mei fora good many years. They were 
men of different countries and very various turns of mind ; but, 
as I mentioned just now, they found no other fundamental idea 
to work with than Dalton’s ; and the result of their Jabours has 
been to confirm the truth of that idea and to extend greatly its 
application. 
One of the results of our endeavours to clas ify substances ac- 
cording to their natural resemblances has been the discovery of 
distinct family relationships among atoms, each family being 
distinguished by definite characteristics. Now among the pro- 
perties which thus characterise particular families of atoms, 
there is one of which the knowledge gradually worked out by the 
labours of an immense number of investigators must be admitted 
to constitute one of the most important additions ever made to 
our knowledge of these'little masses, 
I will endeavour to explain it to you by a simple example. 
An atom of chlorine is able to combine with one atom of hydro- 
gen or one atom of potassium ; but it cannot combine with two 
atoms. Anatom of oxygen, on the other hand, can combine 
with two atoms of hydrogen or with two atoms of potassium, or 
with one atom of hydrogen and one of potassium ; but we can- 
not get it in combination with one atom of hydrogen or of potas- 
sium solely. 
Again, an a‘'om of nitrogen is known in combination with 
three atoms of hydrogen; while an atom af carbon combines 
with four of hydrogen. Other atoms are classified, from their 
resemblance to these respectively, as Monads, Dyads, Triads, 
Tetrads, &c. 
The combining value which we thus recognise in the atoms of 
these several classes has led us naturally to a consideration of 
the order in which atoms are arranged in a molecule. ‘Thus, in 
the compound of oxygen with hydrogen and potassium, each of 
these latter atoms is directly combined with the oxygen, and the 
atom of oxygen serves as a connecting link between them, Ily- 
drogen and potassium have never been found capable of uniting 
directly with one another; but when both combined with one 
atom of oxygen they are in what may be called indirect com- 
bination with one another through the medium of that oxygen. 
One of the great difficulties of chemistry some few years ago 
was to explain the constitution of isomeric compounds, those 
compounds whose molecules contain atoms of like kinds and in 
equal numbers, but which differ from one another in their pro- 
perties. Thus a molecule of common ether contains four atoms 
of carbon, ten atoms of hydroger, and one of oxygen, Butylic 
alcohol, avery different substance, has prec’sely the same com- 
Fosition. We now know that in the former the atom of oxygen 
is in the middle ofa chain of carbon atoms, whereas in the latter 
it is at one end of that chain. You might fancy it impossible 
to decide upon anything like consistent evidence such questions 
as this; but I can assure you that the atomic theory, as now 
used by chemists, leads frequently to conclusions of this kind, 
which are confirmed by independent observers, and command 
general assent. ‘That these conclusions are, as far as they go, 
true descriptions of natural phenomena is shown by the fact that 
each of them serves in its turn as a stepping-stone to further 
discoveries, 
One other extension of our knowledge of atoms I must briefly 
mention, one which has as yet received but little attention, yet 
which will, I venture to think, be found serviceable in the study 
of the forces which bring about chemical change. 
The original view of the constitution of molecules was statical; 
and chemists only took cognizance of those changes of place 
among their atoms which result in the disappearance of the 
molecules employed, and the appearance of new moleclules 
formed by their reaction on one another. Thus, when a solu- 
tion of common salt (sodic chloride) is mixed with a solution of 
silver nitrate, it is well-known that the metallic atoms in these 
respective compounds change places with one another, forming 
silver chloride and sodic nitrate ; for the silver chloride soon 
settles to the bottom of the solution in the form of an insoluble 
powder, while the other product remains dissolved in the liquid. 
But as long as the solution of salt remained undecomposed, each 
little molecule in it was supposed to be chemically at rest. A 
particular atom of sodium which was combined with an atom of 
chlorine was supposed to remain steadily fixed to it. When 
this inactive solution was mixed with the similarly inactive solu- 
tion of silver nitrate, the interchange of atoms known to take 
place between their respective molecules was nominally ex- 
plained by the force of predisposing affinity. It was, in fact, 
supposed that the properties of the new compounds existed and 
produced effects before the compounds themselves had been 
formed. 
I had occasion to point out a good many years ago that mole- 
cules which appear to be chemically at rest are acting on one 
another, when in suitable conditions, in the same kind of way as 
those which are manifestly in a state of chemical change—that 
for instance, the molecules of liquid sodic chloride exchange 
sodium atoms with one another, forming new molecules of the 
same compound undistinguishable from the first, so that, in an 
aggregate of like molecules, the apparent atomic rest is the result 
of the interchange of like atoms between contiguous molecules. 
Such exchanges of atoms take place not only between mole- 
cules of identical composition, but also between contiguous 
molecules containing different elements. For instance, in a 
mixture of sodic chloride and potassic iodide an interchange of 
metallic atoms takes place, forming potassic chloride and sodic 
iodide. The result of the exchange in such a case is to form a 
couple of new molecules different from the original couple. But 
these products are subject to the same general law of atomic 
changes, and their action on one another reproduces a couple of 
molecules of the materials. 
Thus a liquid mixture formed from two compounds contains 
molecules of four kinds, which we may describe as the two 
materials and the two products. The materials are reacting on 
one another, forming the products ; ond these products are, in 
their turn, reacting on one another, reproducing the materials. 
If one of the products of atomic exchange between two mole- 
cules is a solid while the other remains liquid (as when sodic 
chloride is. mixed with silver nitrate), or if one is gaseous while 
the other remains liquid, so that the molecules of the one kind 
cannot react on those of the other kind and reproduce the mate- 
rials, then the continued reaction of the materials on one another 
leads to their complete mutual decomposition. Such complete 
mutual decomposition of two salts takes place whenever they 
react on one another under such conditions that the products’ 
cannot react on one another and reproduce the materials ; whereas 
partial decomposition takes place whenever the materials form a 
homogeneous mixture with the products. 
Now, if in any such homogeneous mixture more exchanges of 
atoms take place between the materials than between the pro- 
ducts, the number of molecules of the products is increased, 
because more of them are being made than unmade ; and reci« 
procally, if more exchanges of atoms take place between the 
products than between the materials, the number of molecules — 
of the materials is increased. ‘The mixture remains of constant 
composition when there are in the unit of time as many decom- 
posing changes as reproducing changes. 
Suppose that we were to determine by experiment the propor- 
tion between the number of molecules of the materials, and the 
number of molzcules of the products, in a mixture the compo- 
si ion of which remains constant, and that we found, for instance, 
twice as many of materials as of products; what would this 
mean? Why, if every two couples of materials only effect in’ 
the unit of time as many exchanges as every one couple of pro-- 
