APPENDIX. 
1 30 
Before sex was invented, as it were, we may suppose the 
various ids in the germ plasm of an individual to have been all 
exactly alike ; but the individuals were not all alike : there were 
individual differences as there are now, and these personal 
differences between individuals corresponded to differences in 
the constitution of the ids of their germ plasms. When, however, 
two different individuals fused their germ plasms into one, the 
resulting germ plasm must have consisted of a mixture of ids of 
two different kinds. These two different kinds of ids, passing into 
the cells of the resulting daughter cells or daughter individuals, 
caused them to have a character intermediate between the 
characters of the two parents. When the second generation 
came to perform amphimixis with a generation formed from the 
fusion of two other individuals, the germ plasm of the third 
generation would presumably consist of ids of four different kinds, 
that of the fourth of eight, and so on until the ids of the germ 
plasm were all different from each other. But when amphimixis 
occurs, we might suppose that the number of ids in the germ 
plasm would be doubled and the germ cells consequently increased 
in bulk. This result is prevented, however, by the reducing 
divisions of which I have before spoken. It is obvious that if 1 
wish to mix a glass of ginger-beer with a glass of stout, 1 shall 
require a glass double the size for the mixture. If, however, I 
throw away half the stout and half the ginger-beer before mixing 
them, an ordinary glass will suffice. What Nature does is to 
throw away half the ginger-beer (i.e., the female germ plasm), 
but she provides an extra glass for the stout, and tries -not to 
waste any of it — but she divides it all the same, and throws a good 
deal of it away eventually. 
It is significant that in these reducing divisions, both male 
and female, the idants do not become split longitudinally, and 
their resulting halves become distributed equally amongst the two 
daughter nuclei, as in ordinary nuclear division — but one-half of 
the entire number of rods passes into one daughter nucleus and 
the other half into the other. 
The result of all this is that every fresh amphimixis results 
in a different combination of ids, and that is why no two children 
of the same parents, whether animal or vegetable, are ever 
exactly alike. Or rather I should say that is the reason why the 
differences are often so great between individuals of the same 
family. There must be variation, independent of amphimixis, 
otherwise amphimixis itself could not cause variation. What, 
then, is the primary cause of variation ? This is a question of 
extreme importance and interest to us as students of British Ferns. 
It has been very largely assumed that differences of environ- 
ment directly caused modifications of structure, and that these 
