Nov., 1890. 
ORGANIC DEATH. 
247 
fraction of the number which are so prematurely killed. Of 
the innumerable seeds produced by plants not one in 10,000 
will germinate, and of those which do germinate not one in a 
thousand will reach maturity. Nevertheless, those which do 
reach maturity will inevitably die of themselves at last, 
So also with a species. It originates from some germ 
which varies abnormally from its parent. This germ fulfils 
its destiny, produces seeds, and dies. One of these seeds 
varies in the same direction as the first, but a little more 
widely. It also produces seeds and dies, and this process 
may go on until a point is reached at which variation in that 
direction ceases. A new species is then established, which 
may go on reproducing itself with little variation for a long 
period. But though the normal variation may be little, it is 
sure and inevitable, and will tend alwavs in the direction of 
the next of the four stages beyond that from which it started. 
If its first germ was the seed of a Conifer, there will be a 
normal tendency to vary in the direction of broader foliage, 
and the change of constitution which that implies. This 
tendency may not become visible for very many generations, 
but in perhaps ten or twenty thousand years a perceptible 
change will have taken place—a change, however, which no 
living man would naturally perceive, and, with the prevailing 
views of scientists, if the earlier form were found as a fossil 
it would not be accepted as belonging to the same species. 
I shall here no doubt be met by the objection that a 
species is not a definite entity, but only an arbitrarily 
separated portion of a continuous current of change. 
This is the prevailing view at the present time, but it can 
hardly be said to be proved. Is not the chain of life really 
a “ chain ” of many links, varying in size, but, though in 
unbroken contact, separated by nodes, these nodes represent¬ 
ing those points and periods of special change which I have 
elsewhere called “critical epochs ?” 
Is not a true species a variety which has passed a “critical 
epoch,” beyond which it cannot again revert, and which 
within its narrow limits will expand, pass through the 
normal organic phases and die, as a species , the branching 
chain being continued by some of its variations, which also 
oass beyond other critical epochs and become definite links 
themselves. On this hypothesis a genus will be a group of 
such branches of the chain all emanating from one node, 
and passing as a group through the same series of changes to 
ultimate death. 
This law of periodicity does not contravene the law of 
natural selection. Selection is not a force in itself; it is 
