depended were supposed to be variable, and thus distinguished from the invariable characters imposed upon the whole species at its creation. Similarly a natural genus was a group of species, and was marked out from other genera by eternal differences of still greater importance.

We now, however, perceive that the existence of any such groups as genera and species is an arbitrary creation. of the naturalist's mind. All resemblances of plants are natural so far as they express hereditary affinities; but this applies as well to the variations within the species as to the species itself, or to the larger groups. All is a matter of degree. The deeper differences between plants have been produced by the differentiating action of circumstances during millions of years, so that it would naturally require millions of years to undo this result, and prove experimentally that the forms can be approximated again. Sub-species may sometimes have arisen within historical times, and varieties approaching to sub-species may often be produced by the horticulturist in a few years. Such varieties can easily be brought back to their original forms, or, if placed in the original circumstances, will themselves revert to those forms; but according to Darwin's views all forms are capable of unlimited change, and it might possibly be, unlimited reversion if suitable circumstances and sufficient time be granted.

Many fruitless attempts have been made to establish a rigorous criterion of specific and generic difference, so that these classes might have a definite value and rank in all branches of biology. Linnæus adopted the view that the species was to be defined as a distinct creation, saying,1 Species tot numeramus, quot diversæ formæ in principio sunt creatæ;" or again, "Species tot sunt, quot diversas formas ab initio produxit Infinitum Ens; quæ formæ, secundum generationis inditas leges, produxere plures, at sibi semper similes." Of genera he also says, "Genus omne est naturale, in primordio tale creatum." It was a common doctrine added to and essential to that of distinct creation that these species could not produce intermediate and variable forms, so that we find Linnæus obliged by the ascertained existence of hybrids to take a different view 1 Philosophia Botanica (1770), § 157, p. 99. 2 Ibid. § 159, p. 100.

2

in another work; he says,1 "Novas species immo et genera ex copula diversarum specierum in regno vegetabilium oriri primo intuitu paradoxum videtur; interim observationes sic fieri non ita dissuadent." Even supposing in the present day that we could assent to the notion of a certain number of distinct creational acts, this notion would not help us in the theory of classification. Naturalists have never pointed out any method of deciding what are the results of distinct creations, and what are not. As Darwin says, "the definition must not include an element which cannot possibly be ascertained, such as an act of creation." It is, in fact, by investigation of forms and classification that we should ascertain what were distinct creations and what were not; this information would be a result and not a means of classification.

3

2

Agassiz seemed to consider that he had discovered an important principle, to the effect that general plan or structure is the true ground for the discrimination of the great classes of animals, which may be called branches of the animal kingdom. He also thought that genera are definite and natural groups. "Genera," he says, "are most closely allied groups of animals, differing neither in form, nor in complication of structure, but simply in the ultimate structural peculiarities of some of their parts; and this is, I believe, the best definition which can be given of genera." But it is surely apparent that there are endless degrees both of structural peculiarity and of complication of structure. It is impossible to define the amount of structural peculiarity which constitutes the genus as distinguished from the species.

The form which any classification of plants or animals tends to take is that of an unlimited series of subaltern classes. Originally botanists confined themselves for the most part to a small number of such classes. Linnæus adopted Class, Order, Genus, Species, and Variety, and even seemed to think that there was something essentially natural in a five-fold arrangement of groups.5

1 Amanitates Academica (1744), vol. i. p. 70. Quoted in Edinburgh Review, October 1868, vol. cxxviii. pp. 416,

2 Descent of Man, vol. i.

P. 228.

3 Agassiz, Essay on Classification, p. 219. 5 Philosophia Botanica, § 155, p. 98.

417.

4 Ibid. p. 249.

With the progress of botany intermediate and additional groups have gradually been introduced. According to the Laws of Botanical Nomenclature adopted by the International Botanical Congress, held at Paris 1 in August 1867, no less than twenty-one names of classes are recognised-namely, Kingdom, Division, Sub-division, Class, Sub-class, Cohort, Sub-cohort, Order, Sub-order, Tribe, Subtribe, Genus, Sub-genus, Section, Sub-section, Species, Subspecies, Variety, Sub-variety, Variation, Sub-variation. It is allowed by the authors of this scheme, that the rank or degree of importance to be attributed to any of these divisions may vary in a certain degree according to individual. opinion. The only point on which botanists are not allowed discretion is as to the order of the successive sub-divisions; any inversion of the arrangement, such as division of a genus into tribes, or of a tribe into orders, is quite inadmissible. There is no reason to suppose that even the above list is complete and inextensible. The Botanical Congress itself recognised the distinction between variations according as they are Seedlings, Half-breeds, or Lusus Natura. The complication of the inferior classes is increased again by the existence of hybrids, arising from the fertilisation of one species by another deemed a distinct species, nor can we place any limit to the minuteness of discrimination of degrees of breeding short of an actual pedigree of individuals.

It will be evident to the reader that in the remarks upon classification as applied to the Natural Sciences, given in this and the preceding sections, I have not in the least attempted to treat the subject in a manner adequate to its extent and importance. A volume would be insufficient for tracing out the principles of scientific method specially applicable to these branches of science. What more I may be able to say upon the subject will be better said, if ever, when I am able to take up the closelyconnected subjects of Scientific Nomenclature, Terminology, and Descriptive Representation. In the meantime, I have wished to show, in a negative point of view, that natural classification in the animal and vegetable kingdoms is a special problem, and that the particular methods and

1 Laws of Botanical Nomenclature, by Alphonse Decandolle, trans Jated from the French, 1868, p. 19.

difficulties to which it gives rise are not those common to all cases of classification, as so many physicists have supposed. Genealogical resemblances are only a special case of resemblances in general.

Unique or Exceptional Objects.

In framing a system of classification in almost any branch of science, we must expect to meet with unique or peculiar objects, which stand alone, having comparatively few analogies with other objects. They may also be said to be sui generis, each unique object forming, as it were, a genus by itself; or they are called nondescript, because from thus standing apart it is difficult to find terms in which to describe their properties. The rings of Saturn, for instance, form a unique object among the celestial bodies. We have indeed considered this and many other instances of unique objects in the preceding chapter on Exceptional Phenomena. Apparent, Singular, and Divergent Exceptions especially, are analogous to unique objects.

In the classification of the elements, Carbon stands apart as a substance entirely unique in its powers of producing compounds. It is considered to be a quadrivalent element, and it obeys all the ordinary laws of chemical combination. Yet it manifests powers of affinity in such an exalted degree that the substances in which it appears are more numerous than all the other compounds known to chemists. Almost the whole of the substances which have been called organic contain carbon, and are probably held together by the carbon atoms, so that many chemists are now inclined to abandon the name Organic Chemistry, and substitute the name Chemistry of the Carbon Compounds. It used to be believed that the production of organic compounds could be effected only by the action of vital force, or of some inexplicable cause involved in the phenomena of life; but it is now found that chemists are able to commence with the elementary materials, pure carbon, hydrogen, and oxygen, and by strictly chemical operations to combine these so as to form complicated organic compounds. So many substances have already been formed that we might be inclined to generalise and infer that all organic compounds might ultimately

be produced without the agency of living beings. Thus the distinction between the organic and the inorganic kingdoms seems to be breaking down, but our wonder at the peculiar powers of carbon must increase at the same time.

In considering generalisation, the law of continuity was applied chiefly to physical properties capable of mathematical treatment. But in the classificatory sciences, also, the same important principle is often beautifully exemplified. Many objects or events seem to be entirely exceptional and abnormal, and in regard to degree or magnitude they may be so termed; but it is often easy to show that they are connected by intermediate links with ordinary cases. In the organic kingdoms there is a common groundwork of similarity running through all classes, but particular actions and processes present themselves conspicuously in particular families and classes. Tenacity of life is most marked in the Rotifera, and some other kinds of microscopic organisms, which can be dried and boiled without loss of life. Reptiles are distinguished by torpidity, and the length of time they can live without food. Birds, on the contrary, exhibit ceaseless activity and high muscular power. The ant is as conspicuous for intelligence and size of brain among insects as the quadrumana and man among vertebrata. Among plants the Leguminosæ are distinguished by a tendency to sleep, folding their leaves at the approach of night. In the genus Mimosa, especially the Mimosa pudica, commonly called the sensitive plant, the same tendency is magnified into an extreme irritability, almost resembling voluntary motion. More or less of the same irritability probably belongs to vegetable forms of every kind, but it is of course to be investigated with special ease in such an extreme case. In the Gymnotus and Torpedo, we find that organic structures can act like galvanic batteries. Are we to suppose that such animals are entirely anomalous exceptions; or may we not justly expect to find less intense manifestations of electric action in all animals?

Some extraordinary differences between the modes of reproduction of animals have been shown to be far less than was at first sight apparent. The lower animals seem to differ entirely from the higher ones in the power of repro