those parts of the structure which determined the habits of life, and the

  general place of each being in the economy of nature, would be of very high

  importance in classification. Nothing can be more false. No one regards

  the external similarity of a mouse to a shrew, of a dugong to a whale, of a

  whale to a fish, as of any importance. These resemblances, though so

  intimately connected with the whole life of the being, are ranked as merely

  'adaptive or analogical characters;' but to the consideration of these

  resemblances we shall have to recur. It may even be given as a general

  rule, that the less any part of the organisation is concerned with special

  habits, the more important it becomes for classification. As an instance:

  Owen, in speaking of the dugong, says, 'The generative organs being those

  which are most remotely related to the habits and food of an animal, I have

  always regarded as affording very clear indications of its true affinities.

  We are least likely in the modifications of these organs to mistake a

  merely adaptive for an essential character.' So with plants, how

  remarkable it is that the organs of vegetation, on which their whole life

  depends, are of little signification, excepting in the first main

  divisions; whereas the organs of reproduction, with their product the seed,

  are of paramount importance!

  We must not, therefore, in classifying, trust to resemblances in parts of

  the organisation, however important they may be for the welfare of the

  being in relation to the outer world. Perhaps from this cause it has

  partly arisen, that almost all naturalists lay the greatest stress on

  resemblances in organs of high vital or physiological importance. No doubt

  this view of the classificatory importance of organs which are important is

  generally, but by no means always, true. But their importance for

  classification, I believe, depends on their greater constancy throughout

  large groups of species; and this constancy depends on such organs having

  generally been subjected to less change in the adaptation of the species to

  their conditions of life. That the mere physiological importance of an

  organ does not determine its classificatory value, is almost shown by the

  one fact, that in allied groups, in which the same organ, as we have every

  reason to suppose, has nearly the same physiological value, its

  classificatory value is widely different. No naturalist can have worked at

  any group without being struck with this fact; and it has been most fully

  acknowledged in the writings of almost every author. It will suffice to

  quote the highest authority, Robert Brown, who in speaking of certain

  organs in the Proteaceae, says their generic importance, 'like that of all

  their parts, not only in this but, as I apprehend, in every natural family,

  is very unequal, and in some cases seems to be entirely lost.' Again in

  another work he says, the genera of the Connaraceae 'differ in having one

  or more ovaria, in the existence or absence of albumen, in the imbricate or

  valvular aestivation. Any one of these characters singly is frequently of

  more than generic importance, though here even when all taken together they

  appear insufficient to separate Cnestis from Connarus.' To give an example

  amongst insects, in one great division of the Hymenoptera, the antennae, as

  Westwood has remarked, are most constant in structure; in another division

  they differ much, and the differences are of quite subordinate value in

  classification; yet no one probably will say that the antennae in these two

  divisions of the same order are of unequal physiological importance. Any

  number of instances could be given of the varying importance for

  classification of the same important organ within the same group of beings.

  Again, no one will say that rudimentary or atrophied organs are of high

  physiological or vital importance; yet, undoubtedly, organs in this

  condition are often of high value in classification. No one will dispute

  that the rudimentary teeth in the upper jaws of young ruminants, and

  certain rudimentary bones of the leg, are highly serviceable in exhibiting

  the close affinity between Ruminants and Pachyderms. Robert Brown has

  strongly insisted on the fact that the rudimentary florets are of the

  highest importance in the classification of the Grasses.

  Numerous instances could be given of characters derived from parts which

  must be considered of very trifling physiological importance, but which are

  universally admitted as highly serviceable in the definition of whole

  groups. For instance, whether or not there is an open passage from the

  nostrils to the mouth, the only character, according to Owen, which

  absolutely distinguishes fishes and reptiles--the inflection of the angle

  of the jaws in Marsupials--the manner in which the wings of insects are

  folded--mere colour in certain Algae--mere pubescence on parts of the

  flower in grasses--the nature of the dermal covering, as hair or feathers,

  in the Vertebrata. If the Ornithorhynchus had been covered with feathers

  instead of hair, this external and trifling character would, I think, have

  been considered by naturalists as important an aid in determining the

  degree of affinity of this strange creature to birds and reptiles, as an

  approach in structure in any one internal and important organ.

  The importance, for classification, of trifling characters, mainly depends

  on their being correlated with several other characters of more or less

  importance. The value indeed of an aggregate of characters is very evident

  in natural history. Hence, as has often been remarked, a species may

  depart from its allies in several characters, both of high physiological

  importance and of almost universal prevalence, and yet leave us in no doubt

  where it should be ranked. Hence, also, it has been found, that a

  classification founded on any single character, however important that may

  be, has always failed; for no part of the organisation is universally

  constant. The importance of an aggregate of characters, even when none are

  important, alone explains, I think, that saying of Linnaeus, that the

  characters do not give the genus, but the genus gives the characters; for

  this saying seems founded on an appreciation of many trifling points of

  resemblance, too slight to be defined. Certain plants, belonging to the

  Malpighiaceae, bear perfect and degraded flowers; in the latter, as A. de

  Jussieu has remarked, 'the greater number of the characters proper to the

  species, to the genus, to the family, to the class, disappear, and thus

  laugh at our classification.' But when Aspicarpa produced in France,

  during several years, only degraded flowers, departing so wonderfully in a

  number of the most important points of structure from the proper type of

  the order, yet M. Richard sagaciously saw, as Jussieu observes, that this

  genus should still be retained amongst the Malpighiaceae. This case seems

  to me well to illustrate the spirit with which our classifications are

  sometimes necessarily founded.

  Practically when naturalists are at work, they do not trouble themselves

  about the
physiological value of the characters which they use in defining

  a group, or in allocating any particular species. If they find a character

  nearly uniform, and common to a great number of forms, and not common to

  others, they use it as one of high value; if common to some lesser number,

  they use it as of subordinate value. This principle has been broadly

  confessed by some naturalists to be the true one; and by none more clearly

  than by that excellent botanist, Aug. St. Hilaire. If certain characters

  are always found correlated with others, though no apparent bond of

  connexion can be discovered between them, especial value is set on them.

  As in most groups of animals, important organs, such as those for

  propelling the blood, or for aerating it, or those for propagating the

  race, are found nearly uniform, they are considered as highly serviceable

  in classification; but in some groups of animals all these, the most

  important vital organs, are found to offer characters of quite subordinate

  value.

  We can see why characters derived from the embryo should be of equal

  importance with those derived from the adult, for our classifications of

  course include all ages of each species. But it is by no means obvious, on

  the ordinary view, why the structure of the embryo should be more important

  for this purpose than that of the adult, which alone plays its full part in

  the economy of nature. Yet it has been strongly urged by those great

  naturalists, Milne Edwards and Agassiz, that embryonic characters are the

  most important of any in the classification of animals; and this doctrine

  has very generally been admitted as true. The same fact holds good with

  flowering plants, of which the two main divisions have been founded on

  characters derived from the embryo,--on the number and position of the

  embryonic leaves or cotyledons, and on the mode of development of the

  plumule and radicle. In our discussion on embryology, we shall see why

  such characters are so valuable, on the view of classification tacitly

  including the idea of descent.

  Our classifications are often plainly influenced by chains of affinities.

  Nothing can be easier than to define a number of characters common to all

  birds; but in the case of crustaceans, such definition has hitherto been

  found impossible. There are crustaceans at the opposite ends of the

  series, which have hardly a character in common; yet the species at both

  ends, from being plainly allied to others, and these to others, and so

  onwards, can be recognised as unequivocally belonging to this, and to no

  other class of the Articulata.

  Geographical distribution has often been used, though perhaps not quite

  logically, in classification, more especially in very large groups of

  closely allied forms. Temminck insists on the utility or even necessity of

  this practice in certain groups of birds; and it has been followed by

  several entomologists and botanists.

  Finally, with respect to the comparative value of the various groups of

  species, such as orders, sub-orders, families, sub-families, and genera,

  they seem to be, at least at present, almost arbitrary. Several of the

  best botanists, such as Mr. Bentham and others, have strongly insisted on

  their arbitrary value. Instances could be given amongst plants and

  insects, of a group of forms, first ranked by practised naturalists as only

  a genus, and then raised to the rank of a sub-family or family; and this

  has been done, not because further research has detected important

  structural differences, at first overlooked, but because numerous allied

  species, with slightly different grades of difference, have been

  subsequently discovered.

  All the foregoing rules and aids and difficulties in classification are

  explained, if I do not greatly deceive myself, on the view that the natural

  system is founded on descent with modification; that the characters which

  naturalists consider as showing true affinity between any two or more

  species, are those which have been inherited from a common parent, and, in

  so far, all true classification is genealogical; that community of descent

  is the hidden bond which naturalists have been unconsciously seeking, and

  not some unknown plan of creation, or the enunciation of general

  propositions, and the mere putting together and separating objects more or

  less alike.

  But I must explain my meaning more fully. I believe that the arrangement

  of the groups within each class, in due subordination and relation to the

  other groups, must be strictly genealogical in order to be natural; but

  that the amount of difference in the several branches or groups, though

  allied in the same degree in blood to their common progenitor, may differ

  greatly, being due to the different degrees of modification which they have

  undergone; and this is expressed by the forms being ranked under different

  genera, families, sections, or orders. The reader will best understand

  what is meant, if he will take the trouble of referring to the diagram in

  the fourth chapter. We will suppose the letters A to L to represent allied

  genera, which lived during the Silurian epoch, and these have descended

  from a species which existed at an unknown anterior period. Species of

  three of these genera (A, F, and I) have transmitted modified descendants

  to the present day, represented by the fifteen genera (a14 to z14) on the

  uppermost horizontal line. Now all these modified descendants from a

  single species, are represented as related in blood or descent to the same

  degree; they may metaphorically be called cousins to the same millionth

  degree; yet they differ widely and in different degrees from each other.

  The forms descended from A, now broken up into two or three families,

  constitute a distinct order from those descended from I, also broken up

  into two families. Nor can the existing species, descended from A, be

  ranked in the same genus with the parent A; or those from I, with the

  parent I. But the existing genus F14 may be supposed to have been but

  slightly modified; and it will then rank with the parent-genus F; just as

  some few still living organic beings belong to Silurian genera. So that

  the amount or value of the differences between organic beings all related

  to each other in the same degree in blood, has come to be widely different.

  Nevertheless their genealogical arrangement remains strictly true, not only

  at the present time, but at each successive period of descent. All the

  modified descendants from A will have inherited something in common from

  their common parent, as will all the descendants from I; so will it be with

  each subordinate branch of descendants, at each successive period. If,

  however, we choose to suppose that any of the descendants of A or of I have

  been so much modified as to have more or less completely lost traces of

  their parentage, in this case, their places in a natural classification

  will have been more or less completely lost,--as sometimes seems to have

  occurred with existing organisms. All the descendants of the genus F,

  along its whole line of descent, are suppose
d to have been but little

  modified, and they yet form a single genus. But this genus, though much

  isolated, will still occupy its proper intermediate position; for F

  originally was intermediate in character between A and I, and the several

  genera descended from these two genera will have inherited to a certain

  extent their characters. This natural arrangement is shown, as far as is

  possible on paper, in the diagram, but in much too simple a manner. If a

  branching diagram had not been used, and only the names of the groups had

  been written in a linear series, it would have been still less possible to

  have given a natural arrangement; and it is notoriously not possible to

  represent in a series, on a flat surface, the affinities which we discover

  in nature amongst the beings of the same group. Thus, on the view which I

  hold, the natural system is genealogical in its arrangement, like a

  pedigree; but the degrees of modification which the different groups have

  undergone, have to be expressed by ranking them under different so-called

  genera, sub-families, families, sections, orders, and classes.

  It may be worth while to illustrate this view of classification, by taking

  the case of languages. If we possessed a perfect pedigree of mankind, a

  genealogical arrangement of the races of man would afford the best

  classification of the various languages now spoken throughout the world;

  and if all extinct languages, and all intermediate and slowly changing

  dialects, had to be included, such an arrangement would, I think, be the

  only possible one. Yet it might be that some very ancient language had

  altered little, and had given rise to few new languages, whilst others

  (owing to the spreading and subsequent isolation and states of civilisation

  of the several races, descended from a common race) had altered much, and

  had given rise to many new languages and dialects. The various degrees of

  difference in the languages from the same stock, would have to be expressed

  by groups subordinate to groups; but the proper or even only possible

  arrangement would still be genealogical; and this would be strictly

  natural, as it would connect together all languages, extinct and modern, by

  the closest affinities, and would give the filiation and origin of each

  tongue.

  In confirmation of this view, let us glance at the classification of

  varieties, which are believed or known to have descended from one species.

  These are grouped under species, with sub-varieties under varieties; and

  with our domestic productions, several other grades of difference are

  requisite, as we have seen with pigeons. The origin of the existence of

  groups subordinate to groups, is the same with varieties as with species,

  namely, closeness of descent with various degrees of modification. Nearly

  the same rules are followed in classifying varieties, as with species.

  Authors have insisted on the necessity of classing varieties on a natural

  instead of an artificial system; we are cautioned, for instance, not to

  class two varieties of the pine-apple together, merely because their fruit,

  though the most important part, happens to be nearly identical; no one puts

  the swedish and common turnips together, though the esculent and thickened

  stems are so similar. Whatever part is found to be most constant, is used

  in classing varieties: thus the great agriculturist Marshall says the

  horns are very useful for this purpose with cattle, because they are less

  variable than the shape or colour of the body, &c.; whereas with sheep the

  horns are much less serviceable, because less constant. In classing

  varieties, I apprehend if we had a real pedigree, a genealogical

  classification would be universally preferred; and it has been attempted by

  some authors. For we might feel sure, whether there had been more or less

  modification, the principle of inheritance would keep the forms together