modes of heredity have been defined by Darwin, and yet more systematically by Haeckel, as "laws of inheritance," and corroborated by them by a profusion of examples. If heredity may be termed the conservative element in the life of species, we may also speak in particular of a conservative heredity, by which the old, long-established characteristics and peculiarities are transferred. The more stubbornly a character is transmitted, or, what amounts to the same, the greater the number of families, genera, and species, over which a character is extended, the more ancient must it be considered, the earlier did it appear in the ancestral stock. In most cases, this conservative heredity occurs in an unbroken succession of generations, an observation on which it is needless to enlarge, as it may be daily made by every one. But conservative heredity may likewise display itself intermittently, either when merely individual characters of the ancestors reappear after lying dormant for one, several, or many generations,-a phenomenon designated as Atavism, or reversion,-or when the species is composed of a regular alternation of variously constituted generations and individuals. This particular sort of reversion is termed Alternate Generation, or Heterogenesis.

No one is surprised if children exhibit the bodily or mental features of their grand-parents which were suspended in the parents. But most frequent and striking is the atavism of domestic animals and cultivated plants, a stubborn antagonist to breeders. Of no domestic animal is the aboriginal stock known with such approximate certainty as that of the pigeon. Now there

are races of pigeons purely bred for several centuries, and in colour and shape transformed into new creatures, which yet from time to time spontaneously, or by crossing with other conspicuous races, produce birds which, in colouring and characteristic pencilings of black bars on wings and tail, resemble the wild rockpigeon.

"I paired," says Darwin, "a mongrel female. barb-fantail with a mongrel male barb-spot, neither of which mongrels had the least blue about them. Let it be remembered that blue barbs are excessively rare; that spots, as has been already stated, were perfectly characterized in the year 1676, and breed perfectly true; this likewise is the case with white fantails, so much so that I have never heard of white fantails showing any other colour. Nevertheless, the offspring from the above two mongrels was of exactly the same blue tint over the whole back and wings as that of the wild rock-pigeon of the Shetland Islands; the double black wing-bars were equally conspicuous; the tail was exactly alike in all its characters, and the croup was pure white."

58

Another reversion frequently to be observed is the striping of the feral domestic cat of Europe, in which it resembles the wild-cat so closely as to be scarcely distinguishable. Darwin has collected evidence from which we may infer that the wild ancestral stock of the horse was striped, and this evidence includes the appearance of striped individuals. But yet another strange phenomenon in horses may be interpreted by atavism. Foals are occasionally born with supernumerary toes. This "monstrosity" can be explained only by

reversion to the three-toed historical ancestors of the present genus. These vouchers are sufficient.

All the phenomena of artificial breeding, as well as natural selection, serve to show that not only the characters descended from past ages, but also those subsequently and most recently acquired, may be transmitted to posterity. This is progressive heredity. Without it, improvement and progress would be impossible; and its own possibility is the direct result of the nature of reproduction. The newer a useful modification, the less has it hitherto been able to place itself in correlation with the entire organism, the less is the reproductive system as yet affected by it; the more uncertain and fluctuating, therefore, is the transfer by propagation; breeding, or natural selection, is requisite to convert the potentiality of progress into a fact, and gradually to enrol this fact among the conservative inheritances.

Progressive heredity is naturally more complex where the sexes are separate, where sexual selection asserts its rights, and the advantages of one sex are fostered by the taste of the other, and are then either transferred exclusively to the sex benefited by its secondary characters, or turned to the profit of the whole species. As a rule, the males are endowed with these advantages, and have transmitted them incompletely to the females. We will explain ourselves by a single example. In the order of insects termed Orthoptera (or straight-winged), the males, by rubbing their wing-covers together, or by stroking them with the lower portion of their hind legs, are able to make a music attractive to the females. Von Graber, a distinguished modern entomologist, has shown 59 that the teeth of the stridulating instruments of

these animals are merely modified hairs; that their construction may be explained by their use; and that in all probability they have been perfected by sexual selection, the best and loudest musicians being the most favoured wooers. With one single exception, the females of the Orthoptera are dumb, but many possess traces of the stridulating apparatus peculiar to the males. Contrary to the older opinion, that it was merely a case of transmission emanating from the males, Graber has made it "more than probable that the resonant nervures of the females-of the stridulating Ephippigera vitium-have been gradually developed independently of the males, but in the same manner." In other cases, on the contrary, the feebly developed nervures of the females, unfit to produce audible stridulations, seem to be an inheritance from the males.

Heredity at corresponding periods of life is a wellknown phenomenon. The tendency to disease is transmitted from the father or the mother to the child to break out at the age at which they suffered. Generation after generation, the milk teeth make room for the permanent teeth at a corresponding time. But all special cases are mere results of the general law of development, by which in the individual characters appear in the sequence in which they were historically acquired and became susceptible of transmission. Heredity, at a definite age after the period at which we consider actual development to be complete, is after all only a continuation of the embryonic development, beginning with fission, germ and ovum, of which the ninth chapter will teach us the signification. In this development of the individual, or ontogenesis, as will be shown below in

more detail, processes are frequently abridged or totally omitted which once, while they were being acquired and after they had been established, occupied a longer period, but in the course of selection either became of less importance to the individual, or preserved a physiological value only as phases of transition.

The second great class of characters, namely, those which have been newly acquired and depend on adaptation, pre-suppose the mutability of the organism. This is a fundamental phenomenon of organic`bodies. It is inherent in the minutest morphological constituents, in protoplasm, and in cells, and in the morphological elements evolved from them, the pervading and determining individual life of which results in the collective life of the creature. The organic morphological element is in a state of saturation; it is continually imbibing and emitting, and its stability is therefore constantly dependent on the supply of material for its functions. For nutrition, which generally and wholly determines the external appearance and the nature of the individual, is accomplished by the innumerable cells and their derivatives. Every fluctuation of supply in any part of the organism, nay, in any point in the surface of a microscopic reef-builder, must necessarily involve a modification of textural parts, or of integrated textural groups or organs.

Mutability is thus a character resulting from the intrinsic nature of organism, and dependent on external conditions which determine quantity and form, as well as the development and transformation of the elementary constituents, or their abortion and retrogression. These effects may be exhibited in a polypestem, which as a whole represents the individual, in its