more highly organized the species is, the more dependent it is on the frequent recurrence of sexual reproduction in the genetic cycle. At least, we find that in the lowest forms there may be a very prolonged pullulation of gemmæ, the sexual act recurring only at distant intervals, and in some cases not being as yet positively known to occur at all; while in the higher animals we meet with no obvious phenomena at all of the nature of gemmation. In those of the lower species in which both modes of propagation are well-marked features, we find that they have a tendency to succeed each other in a regular order, with corresponding differences in the immediate progeny, to which the term of alternation of generations has been applied; and this expression, though open to some objections, has come into very general use.

A complete parallelism, however, cannot be maintained for all the cases that go under this name; and as I am not aware of any systematic analysis having been made, to determine the nature of the differences, it is my object on the present occasion to bring forward certain distinctions which have impressed themselves on my mind as of fundamental importance, depending principally on the period in the life-history of the species, at which a process of gemmation is interpolated in the genetic cycle.

The gemmation sometimes occurs just before, and is, as it were, ancillary to sexual reproduction-sometimes it occurs after it, when it is subservient rather to the progress of development. In the former case, what may, on the whole, be considered as the most typical of the diverse forms belonging to the species, is still defective in having no proper organs of reproduction—a function which is vicariously performed by a set of gemmæ detached from it. The original stock is really neuter; but true sexes appear in these buds, after they have been transformed by a process of development into isolated zooids or phytoids. They may be considered as a highly individualized form of those organs which were wanting in the parent stock. Such organs constitute, at least, the essential part of their economy; and although, along with them, there may be present also others, more or less fully developed, for

discharging functions, such as alimentation and locomotion, required by their status as free zooids, yet their great office is reproduction, and this end effected, their life speedily comes to a close. In this they contrast strikingly with the stock from which they were derived; for it is endowed with much greater permanence of life, frequently detaching during its period of vigour many successive swarms of sexual zooids, just as among the higher animals the same parent may develope many successive broods of young.

On the other hand, when the budding process occurs in the course of development, the gemmæ are detached from the immediate product of impregnation, while it is still in a rudimentary condition, comparable to the first stage in the evolution of the ovum of the higher animals. The germparent never itself attains to the full development of the species, but remains the whole term of its brief existence in a rudimentary state; but the progeny, which it buds off, acquire, in due course, the typical form, or at least give origin, mediately or immediately, to others which do so.

These two kinds of zooids, however, though differing so widely in their relations and structure-in the one case the primary products of impregnation, the precursors of the perfect form, and without sexual characters, in the other derivative and with distinct sexes-have yet this one point in common, that the great end of their existence is the multiplication of the race-an end to which the nutritive and animal functions are always subordinated. Such, indeed, is the occasional degradation or non-development of structure, that some zooids of both kinds might readily pass for mere eggsacs or proliferous cysts.

§ 4. Protomorphic Alternation.

For the better distinction of these varieties of alternation, and for the purpose of bringing out more clearly what I consider to be their points of correspondence with phenomena occurring in the higher animals, I have found it convenient to divide the life-history of an organic being into three stages,

all of which come out prominently in one form of alternation or other, while, as I shall presently endeavour to show, they are covertly represented even in those species, in which no phenomena of alternation are recognised. The first, or what I term the Protomorphic stage, is that which intervenes between the fecundation of the germ and the first appearance of the characteristic or typical organisation of the species; the second, or Orthomorphic, that which corresponds to the development and full perfection of this organisation; while the third, or Gamomorphic, is that of the formation or maturation of those structures in which the spermatic and germinal elements are generated, in preparation for another act of fecundation, as the commencement of a new genetic cycle.

In one of the forms of alternation just noticed, the interpolation of gemmation takes place in the protomorphic stagethat is, prior to that development, by which the features most characteristic of the species are gradually evolved. Of this we have an example in the case of the Trematode Entozoa, so often referred to by writers on the subject of alternation. In these animals the immediate product of the impregnated ovum is a free zooid, which never rises itself above a rudimentary condition, or acquires sexual organs, but which, by a process of asexual gemmation (monogenesis), ultimately originates others, which do attain to the typical character of the species, in the general organisation, and commonly also in the sexual relations, and then propagate in the manner of the higher orders.

For another illustration, we may turn to an allied family, the Cystic Entozoa, now known to be merely rudimentary forms of Cestoid worms. Their transformation into the latter is their most notable change, but, prior to it, they present us with a series of successive forms, all referable to the cystic phase. The typical form is the " Tænia-head," which is not the immediate product of impregnation, but is derived as a gemma from the primary cyst, into which the contents of the ovum are first developed. With great differences of detail, this general relation is to be traced in all the species. Thus, in the Echinococcus hominis, a vesicular mass is formed from

the primary cyst, by the pullulation from its interior of secondary and tertiary structures of a like kind. Numerous gemmæ are developed from the last-formed cysts, having the general characters of "Tania-heads." In another species or variety of Echinococcus, similar" Tænia-heads" are formed, in connection with vesicles budded off from the interior of the primary cyst, without the intermediate pullulation of other cysts. In Conurus, also, there is a formation of a multitude of "Tania-heads" from the original cyst, only they are budded off from a special thickening of the lining membrane -not as in Echinococcus, from its whole interior.

The Echinodermata might also be cited in illustration, though they differ from the generality of cases of alternation, in the primary form budding off but a single secondary one.

As an example of protomorphic alternation in the vegetable kingdom, the case of the mosses may be referred to. Impregnation is now admitted to be as necessary a step in the reproduction of these as of any phanerogamic plants, but it is not the immediate precursor of the formation of an embryo. In mosses, the germinal element is represented by the central cell of the archegonium; and this, when fecundated by the spermatic filaments contained in the cellules of the antheridium, developes by endogenous formation a whole mass of cells, which, by a process of transformation in the course of growth, assumes eventually the form of the theca or capsule, with seta, calyptra, operculum, peristome, columella, and internal mass of dust-like spores. The spores in germination give rise to confervoid threads, which, after ramifying into a mass of tangled filaments-the protonema-send up here and there a leafy axis, bearing eventually, like the original one, antheridia and archegonia. In thus regarding the case of the mosses as parallel to that of the Trematoda, and illustrative of protomorphic alternation, of course I look upon the nascent axis as the true embryo, and the fully-developed moss as the typical form, and regard both the theca and the protonema as intermediate forms, no more represented in the higher plants, than the gregariniform zooids of distoma are in the higher grades of animal life.

§ 5. Gamomorphic Alternation.

The other form of alternation, before referred to, is that in which the process of gemmation is interpolated in what has been here termed the gamomorphic stage, i.e., after the general acquisition of the typical conformation of the species, and in connection with the development of the organs which form the sexual clements. To this head I refer the alternation of polype and medusa forms, which is so common a feature in the life-history of the hydraform zoophytes, the normal or typical form being assumed to be that of the polype, and the medusa form being regarded simply as a highly individualized generative organ detached from the system of the polype. This view readily enough commends itself to our judgment in those species of Laomedea, &c., in which the medusiform zooids have so much the character of mere generative appendages, while the zoophyte condition stands prominently forward as the typical or orthomorphic phase, being represented by structures of greater permanence than the brood of minute and rudimentary medusoids which they throw off, and occasionally attaining considerable dimensions by the repeated pullulation of new polypes. But in the case of the Hood-eyed Medusæ, the primâ facie aspect of the case is entirely the other way; for not only are they themselves of much more conspicuous dimensions and elaborate organisation than the medusoids of the compound zoophytes just referred to, but the polype stock from which they spring is of such insignificant proportions, that it generally goes under the name of a larva, the resulting medusæ being regarded as the typical or perfect form of the species. Thus, Professor Owen remarks, "The medusiform ovigerous locomotive or distributive individual of the Coryne and Campanularia geniculata is evidently homologous with the polypiform ovigerous individual, which seems to nurse, as it were, the ova into planulæ' in the Campanularia dichotoma, and the nutritive gemmiparous polypiform individuals in all the compound Radiaries would seem, rather than the oviparous medusiform ones, to manifest the typical form of the species. Superadd, however,