Those who defend the stability of species rather imagine that, with Cuvier, they are entitled to interpret facts in their own favour; whereas they partly remain unconsciously involved in hereditary prejudice, and partly contrive to be deliberately blind to all that evidently contradicts the immutability of species.

Since Linnæus referred to the Creation, he attributed the individuals to a species, of which the pedigree ascended in direct line to the pair which proceeded from the hand of the Creator. Owing to the state of science in general, an examination of this pedigree was totally impossible in his time; and, indeed, with the strict reliance on sacred tradition, it was scarcely necessary. Cuvier, although a very unprejudiced and cool observer, nevertheless radically accepted the Linnæan definition of species. According to him, the species is the aggregate of individuals descending from one another and from common ancestors, and of those who resemble them as strongly as they resemble one. another.21

"In this definition," says Haeckel, "to which the majority have ever since more or less closely adhered, two things are obviously required of an individual as belonging to a species: in the first place, a certain degree of resemblance or approximate similarity of character; and secondly, a kindred connection by the bond of a common descent. In the numerous attempts of later authors to complete the definition, the chief stress is laid sometimes on the genealogical consanguinity of all the individuals, sometimes on morphological uniformity in all essential characters. But it may be generally asserted that in the practical application of

the idea of species, in the discrimination and nomenclature of the individual species, the latter criterion alone has almost always been employed, while the former has been entirely neglected. Later, it is true, the genealogical idea of the common descent of all individuals of each separate species was supplemented by the physiological definition that all the individuals of every species are capable of producing fertile offspring, by intercrossing, whereas sexual intercourse between individuals of different species produces only sterile offspring or none at all. In practice, however, it was considered quite enough if, among a number of extremely similar animals under investigation, uniformity in all essential characters could be established, and no inquiry was made whether these individuals ascribed to the same species were actually of common origin, and capable, by crossing, of producing fertile offspring. The physiological definition was no more applied in the practical discrimination of animal and vegetal species, than was the pre-supposed common descent from a single ancestral pair. On the other hand, two closely allied forms were distinguished without scruple as two different 'good species,' whenever in a number of similar individuals examined a constant difference could be demonstrated, even though of a merely subordinate character. Here, again, no pains were taken to ascertain whether the two different series were not really descended from common ancestors, and were really capable of generating in conjunction only sterile hybrids, if any."

That this radical condemnation of the post-Linnæan manufacture of species is not too severe, is shown by one fact among others; that within the fraternity such

utter discord as to the limitations of species prevailed, and still prevails, that no agreement can be arrived at respecting the basis of the description of species, the "essential characteristics." Although Agassiz lays down the diagnosis of the species, a decision is required in each case as to the mutual relations of the parts, the ornamentation, &c. As in the absence of birds'-nests, snailshells, butterflies, &c., it is impossible, when it comes to the erection of species, to pre-determine what may be the "essential characteristics" of the species they are to form, subjective opinions and arbitrary decisions have full play; and within a certain domain, well known by its forms, there are among the systematizers no two authorities who are agreed as to the number of species into which the material before them should be divided.

The most unbridled license in the manufacture of species prevailed, however, among the Paleontologists during a period when, in the endeavour to fix the subdivisions of geological strata as accurately as possible by means of their organic contents, the separation of species was carried incredibly far, into the most minute and often into individual deviations. A certain mutability of species could not fail to obtrude itself on the most purblind eye; ramifications were made of sub-species, sports of nature, and varieties characterized by "less essential" peculiarities acquired by means of climate and inheritance. There was, however, always a reservation that their crosses with one another and with the main species should produce fertile offspring, whereas towards other species their relations were identical with those of the main species. Of course, in this separation of the species into sub-species, subjective opinion was even

less fettered by tradition and law than in the definition of species. The literature of ornithology during the last forty years could furnish thousands of the strangest examples of the Babel-like confusion which was thus introduced.

There is no question that a great, perhaps the greater, number of organisms now existing are in a condition in which, according to their internal and external relations, they may be characterized by Natural History as so-called species, and for the purpose of recognition ard scientific treatment in general, must needs be so characterized. But this stability, as may be shown both directly and by analogy, is under all circumstances only temporary, and we have whole classes of organisms to which it is impossible, even with the widest reservations, to apply the old idea of species, with its immutability of essential characteristics. If we are able to furnish incontrovertible proofs of the existence of such non-specific groups, the old system and the dogma of species are once for all set aside, and the positive basis of a new doctrine is secured. This evidence is supplied in two directions. Some classes of organisms in their present state vacillate and fluctuate in form, in such a manner that it is utterly impossible to fix the characteristics of species or genus. They are in an extreme grade of mutability, which, in others, has given way to an apparent state of repose. Other series of facts, exhibiting the most obvious mutability of species, are displayed by certain antediluvian groups in the succession of forms called "species.'

Even before the appearance of Darwin's work on the "Origin of Species," Carpenter, in the course of his researches on the Foraminifera, arrived at the con

clusion, proved in special instances, that in this group of low organisms which secrete the most delicate calcareous shells, there could be no question of "species," but only of "series of forms." Forms which the systematists had reduced to different genera and families, he beheld developing themselves from one another. These Foraminifera are, however, so simple in structure, the history of their individual evolution or Ontogenesis is, as yet, so little known; they contribute so little microscopic detail, which might formulate the law of transmutation of species, that the champions of persistency of species might still seek refuge in the assertion that Carpenter's series of forms are mere varieties, and only prove that the true "species" have not yet been found.

We may now turn with advantage to the class of the Spongiada, the importance of which in the question of species I was the first to point out.22 With them, as I summed up my researches, it is not as with the Foraminifera, merely an affair of the general habit of the form, of the variable grouping of the chamber systems; but the variability exists still more specially in the microscopic detail than in the coarser constituents. In the Foraminifera we may speak of microscopic forms, but not properly of microscopic constituents. But in the sponges we discern the transformation of the finer morphological constituents, the rudimentary organs, and we thereby gain an insight into the mutability of the whole. In this respect the calcareous sponges are somewhat differently circumstanced from the rest, and from the silicious sponges in particular. In the former, the variability of the microscopic parts is limited to a smaller circle of forms,