[March, 
124 Thoughts on Chemical Affinity. 
it is also found that when single, as when just released from 
combination with some other element, their chemical vigour 
is far stronger than after they have had time to combine 
among themselves, and form atomic pairs. Such a result 
would necessarily follow if their attractions were magnetic. 
In the free, or nascent, state their magnetic attraction must 
be far more vigorous than after they have formed astatic 
pairs with each other, and thus have become magnetically 
inert. In a case like that of ozone, where three chemical 
atoms of equal strength are conjoined, we would have a 
weak magnetic system, since each magnet must disturb the 
attraction of the other two. One of these oxygen atoms 
might thus readily yield to outer attractions, and the two 
remaining enter into a more vigorous union. But four 
atoms, as in the molecules of phosphorus and arsenic, might 
naturally conjoin in a square or circle, each being astatic to 
its two neighbours. Any even number of atoms must pro- 
duce a more vigorous astatic arrangement than an odd 
number. 
Another peculiar feature of chemical affinity, that of the 
possession of varied bonds of affinity by varied elements, 
may perhaps be equally explicable on the magnetic hypo- 
thesis. It might be that the monad has only one-half the 
magnetic energy of the dyad, one-third that of the triad, 
and so on ; and that, therefore, it would require two monads 
to satisfy the magnetic energy of one dyad, &c. This could 
scarcely be were all the chemical elements prime molecules 
of matter. The magnetic energy of these could scarcely 
possess such exaCl numerical relations. But if the chemical 
elements be really, in great part, more or less complex com- 
pounds of these prime molecules, — if there be but one, cr 
but few, kinds of prime molecules, from whose simpler com- 
binations the chemical elements arise, — then the free energies 
of the latter would necessarily be in strict numerical propor- 
tion. They would comprise sums of equal magnetic units, 
some of these being masked by their astatic combination, 
others directed outwardly, and the element would display as 
many bonds of attraction as it had units of magnetic energy 
directed outwardly. 
There is another chemical condition to which we must 
here refer. This is, that the quantivalence of an element is 
not at all times the same, and that, when it varies, it loses 
or gains two, or a multiple of two, bonds, never — with a few 
possible exceptions — one or three bonds. This faCt is ex- 
plained in ordinary chemical language by saying that the 
free bonds satisfy each other, and that, therefore, they must 
