It is plain that if the character of important parts can be thus changed by changes in other parts, the typical or characteristic form of these parts may be due only partially to heredity.

We see then that the structural complexity of an adult animal is due in part to the formation of structures which are not alive, in part to the direct modifying influence of external conditions of life, and in part to the action of one organ of the body upon another, so that the number of features which are directly inherited is very much less than the number which are constant in and characteristic of the species.

It is impossible for us to state at present how many features must be subtracted from the race characteristics of an animal in order to give us the total number of hereditary congenital characteristics. The observations and experiments which are recorded are few in number, but they are sufficient to show us that, in all the higher animals, very considerable deduction must be made, and we may be sure that the mature animal is vastly more complex than the egg. There is still another limiting circumstance which has not yet been mentioned.

Many of the parts of an organism are due to indefinite multiplication of a single element. The simplest illustrations of this fact are the blood corpuscles of vertebrates and the leaves of plants. It is clearly unnecessary to suppose that each vertetrate ovum contains separate particles for all the blood corpuscles, or that each seed contains separate particles for all the leaves which the plant is to produce. All that is necessary is to assume that it contains particles which are capable of producing a single one of these structures, with a capacity for indefinite multiplication, and that surrounding conditions determine how far, and in what places,

this power of multiplication shall manifest itself. Most of the organs of the body contain great numbers of cells. which are alike both in structure and function, and as it is usually quite impossible to say how far the size of an organ is truly hereditary, and how far it is determined by surrounding conditions, it is, of course, impossible to say to what extent its mature structure is represented in the ovum, but as we know that the size of most organs varies, and may be increased or diminished by external influences, we may be quite certain that the number of independent cells which make up the tissues and organs of a mature organism, is very much greater than the number represented by distinct particles in the ovum.

It is not even necessary to suppose that all classes of cells which are present in the adult are represented in the ovum. In a mammal, for instance, certain epithelial cells become converted into hairs, while others become. converted into glands or other specialized epithelial

structures.

It is not necessary to assume that all of these specializations are represented in the ovum, for we know that ordinary epithelial cells, in a part of the body where no hair is normally developed, may, when inflamed, give rise to hairs. It is therefore quite possible that each epithelial cell may, when excited by the proper influence, tend to become converted into a hair cell. Each cell of the body may possess the tendency to manifest certain properties under certain conditions, and to manifest certain other properties under other conditions, and the descendants of a single cell may thus become modified in several divergent directions, and each modification may be perfectly constant and characteristic of the race without being hereditary; that is, without being represented in the ovum by a particle with the same specialization.

It may seem to some that the assumption that the egg contains particles capable of producing an unspecialized epithelial cell which shall have the power to give rise to all the specialized sorts of epithelial cells, involves just as much complexity of structure as the assumption that each kind of cell is represented in the ovum, but I think an illustration will show that this is not the case.

Training of a certain kind will develop a boy into a good pedestrian, while another sort of training will make him a good shoemaker; but it is surely simpler to assume that he is born with a tendency to develop the characteristics of a shoemaker under the influence of certain conditions, and those of a pedestrian under other conditions, than to assume that he is born with all the peculiarities of both latent in his organization.

The direct modifying influence of surrounding conditions is a subject upon which very much remains to be done, but we know enough about it already to state that many of the constant characteristics of organisms are due to exposure to constant and uniform conditions rather than to heredity. To what extent this is true we are quite unable to determine, but we can be sure that the organization of the ovum is simpler, and in all probability vastly simpler, than that of the developed organism.

After all these deductions are made the number of strictly hereditary features is very great indeed, and the egg of one of the higher animals must be a marvellous structure, for we know that, after all, most of the characteristics of an organism are not due to the influence of its conditions of life, but to the past history of the race; and Darwin has shown us that the successive changes which have resulted in the evolution of any organism do not, usually, owe their existence to the direct modify

ing effect of external influences, but to the natural selection of congenital variations.

The fact that our theory requires us to believe that the egg of one of the higher animals is complex beyond our powers of conception, must not be regarded as an argument against the theory, for we are compelled to believe this in any case. The difference between our theory and other attempts to explain the phenomena of heredity, is that it does what no other hypothesis attempts. It furnishes a simple explanation of the manner in which the ovum has acquired its present complexity.

In the following chapters I shall give some of the reasons for believing that the difference between the functions of the sexual elements which the theory requires does actually exist, but even in the absence of this proof it would be natural to conclude that if race modification could be furthered and aided by the divergent specialization of the functions of the two reproductive elements, natural selection would, in all probability, have acted so as to bring such a specialization about.

We know that the influence of natural selection is constantly exerted to seize upon and perpetuate any tendency to division of labor among the organs and tissues and cells of the body, and it is only natural that the successive stages in the specialization of the sexual elements should have been perpetuated like any other useful specialization.

CHAPTER V.

ON THE OPINION THAT EACH SEX MAY TRANSMIT ANY CHARACTERISTIC WHATEVER.

The argument from hybrids-This argument is inconclusiveThe argument from the homology between the ovum and the male cell-Homology does not involve functional similarity— The argument from the dual personality of each individual; from reversion; and from polymorphism-These phenomena admit of a simpler explanation-Summary of chapter.

The Argument from Hybrids.

According to the view to be presented in this work, the functions of the two sexual elements, in inheritance, are not alike.

The proof of this will be presented further on, when the subject is reached in the logical course of the development of our argument.

Some of the very highest authorities have been led to a view which is directly opposite, and have held that either parent may transmit to the offspring any characteristic whatever. Lest any reader should assume, at the beginning of this book, that the work involves an absurdity, and that my conclusion is already disproved, it seems best to at once examine the reasons for the opposite view. If I can show that these reasons are inconclusive, and that there is and can be no proof for the statement that each sexual element transmits to the offspring every characteristic of the parent, we can then enter into the subject without prejudice, and can wait