is no necessary or constant connection between the fertilization of the egg and the sex of the embryo, and the conclusion which I have reached from the study of these cases and of our scanty information upon the subject from other sources, is that sex is not determined by any constant law; that in certain animals and plants the sex of the embryo is determined by certain conditions, while in other groups it is determined by quite different conditions.

However this may be, it is obvious that since perfect males and perfect females may arise from eggs which are fertilized, and also from eggs which are not fertilized, the necessity for fertilization does not come from the necessity for transmitting to the offspring the organization of each parent.

A review of the opinions and reasoning of various authors shows that there is no good ground for believing that the two reproductive elements play similar parts in heredity and transmit every characteristic of each parent. It is impossible to prove it by the phenomena of crossing, since the only animals which can be made to cross are essentially alike, and differ only in minor points. The homology between the ovum and the male cell is no reason for supposing that their functions are similar. There is no reason for assuming that each sex transmits its entire organization to the offspring, since the latent transmission of secondary sexual characters can be more simply explained by assuming that each embryo inherits, but does not necessarily develop, all the characteristics of its species.

Reversion and alternation of generations admit of a similar explanation, and we may conclude that there is and can be no proof that each sexual element transmits all the characteristics of the parent. There is therefore

no a priori absurdity in the hypothesis that the ovum and the male cell fill different offices. While there is no reason for believing that the functions of these elements are alike, there are many reasons for believing that this is not the case; for example, the almost universal occurrence of differences in form, size, and structure; the possibility of parthenogenesis; the differences between reciprocal hybrids; the fact that the offspring of a male hybrid and a female of a pure species is much more variable than the offspring of a female hybrid by the male of a pure species; and the fact that a part which is more developed or is of more functional importance in the male parent than it is in the female parent, is much more apt to vary in the offspring than a part which is more developed or more important in the mother than it is in the father.

In the absence of all evidence to the contrary I think we may safely conclude from this positive evidence that a division of physiological labor has arisen during the evolution of life, and that the functions of the reproductive elements have became specialized in divergent directions.

The only way to discover the exact nature of this specialization is to study the influence of each element separately, and the comparison of sexual with asexual reproduction is the best available method of doing this, since asexual reproduction is essentially reproduction without a male element, while sexual reproduction. is reproduction with a male element.

Organisms produced from fertilized ova differ from those produced asexually only in their greater tendency to vary, and the hypothesis that the male element has become specialized for the transmission of a tendency to vary naturally suggests itself. Variation is not dependent upon fertilization, for plants produced from buds

vary as well as those born from fertilized seeds, although bud variations are extremely rare as compared with seedling variations.

In any attempt to frame an hypothesis of heredity we must therefore recognize all the following facts: that the two reproductive elements are homologous, and that their functions were originally alike; that the possibility of parthenogenesis, together with many other well ascertained facts, shows that their functions are not alike, in the higher organisms, at present; that their present functions are due to divergent specialization or physiological division of labor; that variation is possible without sexual union, but that the introduction of a male element in reproduction greatly increases the frequency of its occurrence.

Among the unicellular organisms variability is provided for by conjugation, or the fusion of two entire individuals so that the new generation is derived from a compound germ which contains particles to represent all the parts of the body of each parent. In the metazoa and the many celled plants the reproductive bodies are localized and they are single cells, and there must therefore be some mechanism or organization in virtue of which they represent cells from all parts of the body, and thus provide for further variation.

These various considerations have led us to believe that each cell of the organism inherits from its unicellular ancestors the power to throw off cell germs or gemmules; that these germs penetrate to all parts of the body, and that those which thus reach the developing reproductive elements insure variation, in the next generation, in the cells which they represent; that originally the two sexual elements were alike in function; that cach inherited from the fertilized ovum of the pre

ceding generation the power to give rise to a new organism with all the established hereditary characteristics of the race; and that each element also had, by virtue of its contained gemmules, the power to transmit variability.

The existence, in each element, of the power to transmit the hereditary characteristics of the species is obviously superfluous, since the object of sexual union, the transmission of a tendency to vary, would be equally well secured if only one element had the power to transmit the common characteristics of both parents. I therefore believe that, as organisms gradually increased in size, as the number of cells in their bodies grew greater, and as the differentiation and specialization of these cells became more and more marked, one element, the male cell, became adapted for storing up gemmules, and, at the same time, gradually lost its unnecessary and useless power to transmit hereditary characteristics. This process was gradual, and even in the highest animals the power of the male cell to transmit hereditary characters does not seem to be completely lost, although few traces of it remain.

I also suppose that natural selection has acted upon the various cells of the body to restrain them from throwing off unnecessary gemmules, and that this power is exercised only when a change in the surrounding world renders variation necessary.

After framing this hypothesis the next step is to test it by applying it to the various observed phenomena of heredity in order to see how far it explains and interprets them. I have attempted to do this in chapters VI. to X. of this book, and I think we are justified in concluding, as the result of this review, that, while there are many facts which the hypothesis does not explain, they are not of such a character as to directly contradict it, while it

does group and illuminate many classes of facts which are quite inexplicable without it.

The evidence from hybrids seems to be strongly in its favor, and it presents many features which are perfectly simple and natural, according to our view of heredity, although no other explanation of them has ever been offered.

Hybrids and mongrels are highly variable, as we should expect from the fact that many of the cells of their bodies must be placed under unnatural conditions, and must therefore have a tendency to throw off gemmules. Darwin's pangenesis hypothesis accounts for the variability of hybrids, but it does not account for the very remarkable fact that hybrids from forms which have long been cultivated or domesticated are more variiable than those from wild species or varieties, or for the fact that the children of hybrids are more variable than the hybrids themselves.

Our view not only explains the variability of hybrids, but it also accounts for the excessive variability of hybrids from domesticated forms, and of the children of hybrids, for domesticated animals and plants live under unnatural conditions, and they are therefore more prolific of gemmules than wild species, and as the body of a male hybrid is a new thing the cells will be much more likely than those of the pure parent to throw off gemmules.

The fact that variation is due to the male influence, and that the action upon the male parent of unnatural or changed conditions results in the variability of the child, is well shown by crossing the hybrid with the pure species, for when the male hybrid is crossed with a pure female the children are much more variable than those born from a hybrid mother by a pure father.