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 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." With plants how remarkable it is that the organs of vegetation, on which their nutrition and life depend, are of little signification; whereas the organs of reproduction, with their product the seed and embryo, are of paramount importance! So again in formerly discussing certain morphological characters which are not functionally important, we have seen that they are often of the highest service in classification. This depends on their constancy throughout many allied groups; and their constancy chiefly depends on any slight deviations not having been preserved and accumulated by natural selection, which acts only on serviceable characters.

WHAT IS A SPECIES?

"Each kind of animal or plant, that is, each set of forms which in the changes of the ages has diverged tangibly from its neighbors, is called a species. There is no absolute definition for the word species. The word kind represents it exactly in common language, and is just as susceptible to exact definition. The scientific idea of species does not differ materially from the popular notion. A kind of tree or bird or squirrel is a species. Those individuals which agree very closely in structure and function belong to the same species. There is no absolute test, other than the common judgment of men competent to decide. Naturalists recognize certain formal rules as assisting in such a decision. A series of fully intergrading forms, however varied at the extremes, is usually regarded as forming a single species. There are certain recognized effects of climate, of climatic isolation, and of the isolation of domestication. These do not usually make it necessary to regard as distinct species the extreme forms of a series concerned."

'From D. S. Jordan and V. L. Kellogg, Evolution and Animal Life.

"The term 'species' was thus defined by the celebrated botanist De Candolle: 'A species is a collection of all the individuals which resemble each other more than they resemble anything else, which can by mutual fecundation produce fertile individuals, and which reproduce themselves by generation, in such a manner that we may from analogy suppose them all to have sprung from one single individual.' And the zoologist Swainson gives a somewhat similar definition: 'A species, in the usual acceptation of the term, is an animal which, in a state of nature, is distinguished by certain peculiarities of form, size, colour, or other circumstances, from another animal. It propagates, after its kind, individuals perfectly resembling the parent; its peculiarities, therefore, are permanent.'"

[As will have become apparent, the significant assumption underlying classification is that the closest fundamental similarities between animals (or plants) are found in the forms most closely related and that the greatest differences are found in those forms which are unrelated or at best very distantly related. The assumption implies the idea of descent with modification, which is no more nor less than evolution. Using this evolutionary basis, we can arrive at an extremely satisfactory classification both of living and of extinct forms; and there is no other basis of classification that works.

The question might well be asked whether it is possible to test the validity of the assumption that degrees of resemblance vary directly with closeness of blood relationship? Two direct tests of this may be and have been made. The closest of blood relatives possible are individuals that have been derived by the dividing of a single egg. Armadillo quadruplets have been shown to be thus derived, and detailed studies of the closeness of resemblance existing between members of a given set indicate that they are vastly more alike than are the simultaneously born offspring of animals which give birth to several young, but in which each young is derived from a separate egg. If we use the index of correlation to indicate the degree of similarity between individuals we find that ordinary brothers or sisters are only about 50 per cent alike, while armadillo quadruplets are over 90 per cent alike. Identical or duplicate twins in human beings are believed to have an origin from one egg, after the fashion of the armadillo, ' From A. R. Wallace, Darwinism.

See H. H. Newman, The Biology of Twins (1917), University of Chicago Press.

though the proof has not been forthcoming. Everyone is familiar with the remarkable similarity, amounting almost to identity, between such twins. Thus we are able to show that the closest blood relationship known is associated with the closest resemblance. The next degree of resemblance is between members of the same family, brothers, sisters, cousins, etc., and we do not hesitate to explain this resemblance as due to blood relationship. In this we merely accept the known principles of heredity.

The second direct test of the validity of the assumption that degrees of resemblance run parallel with degrees of blood relationship is found in connection with "blood-transfusion tests." This evidence, as presented by Professor Scott, forms the substance of the next chapter.-ED.]

CHAPTER IX

EVIDENCE FROM BLOOD TESTS'

W. B. SCOTT

Here may be conveniently considered the very interesting and significant blood tests which have been made in the last fifteen years by various physiologists and especially by Dr. George H. F. Nuttall, of the University of Cambridge. Though there are several methods of making these tests, the "precipitation method" employed by Dr. Nuttall will be quite sufficient for the ends sought in these lectures. The method and significance of the tests can best be explained by taking as an example human blood, which, of course, has been most extensively and minutely studied, because of its legal importance as well as its scientific interest. Ordinary chemical analysis is unable to determine the differences in blood-composition between various animals, but that there were important differences had long been understood. This was shown by the fact that, in performing the operation for the transfusion of blood, it was not practicable to substitute animal for human blood, since the former might cause serious injury to the patient.

The precipitation method of making blood tests is as follows: Freshly drawn human blood is allowed to coagulate or clot, which it will do in a few minutes, if left standing in a dish, and then the serum is drained away from the clot. Blood-serum is the watery, almost colourless part of the blood, which remains after coagulation. Small quantities of this serum are injected, at intervals of one or two days, into the veins of a rabbit and cause the formation in the rabbit's blood of an anti-body, analogous to the anti-toxin which is produced in the blood of a horse by the injection of diphtheria virus. After the last injection the rabbit is allowed to live for several days and is then killed and bled, the blood is left until it clots and the serum drained off and preserved. The serum obtained thus from a rabbit is called "anti-human" serum and is an exceedingly delicate test for human blood, not only when the latter is fresh, but also when it is in the form of old and dried blood-stains, or even when the blood is

From W. B. Scott, The Theory of Evolution (copyright 1917). Used by special permission of the publishers, The Macmillan Company.

putrid. Stains, for example, are soaked in a very weak solution of common salt and, if necessary, the blood solution is filtered until it is quite limpid and clear. Into the blood solution a few drops of the anti-human serum are conveyed and, if the stains are of human blood, a white precipitate is formed and thrown down, but if the stains are of the blood of some domestic animal, such as a pig, sheep, or fowl, no such reaction follows. In the same manner as above described, we may prepare anti-pig, anti-horse, anti-fowl, etc., etc., sera by injecting the fresh-drawn serum of a pig, horse, fowl, or any other animal into the rabbit, instead of human blood-serum. In some countries, notably in Germany and Austria, this test has already been adopted by the courts of justice and has been found extremely useful in the detection of crime.

Further investigation showed that these blood tests might be employed to determine the degrees of relationship between different animals, for, although a prompt and strong reaction is usually obtained only from the blood of the same species as that from which the original injection into the rabbit was taken, the blood of nearly allied species, such as the horse and donkey, for example, gives a weaker and slower precipitation. By using stronger solutions and allowing more time, quite distant relationships may be brought out. Nuttall and his collaborator, Graham-Smith, made many thousands of such experiments bearing upon the problems of relationship and classification and it is of great significance to note that their highly interesting and important results contain few surprises, but, in almost all cases, merely serve to confirm the conclusions previously reached by other methods, such as comparative anatomy and palaeontology. It will be instructive to quote some of these results, the quotations being taken from "Blood Immunity and Blood Relationship, by G. H. F. Nuttall, including Original Researches by G. L. Graham-Smith and T. S. P. Strangeways," Cambridge, 1904.

"In the absence of palaeontological evidence the question of the interrelationship amongst animals is based upon similarities of structure in existing forms. In judging of these similarities, the subjective element may largely enter." "The very interesting observations upon the eye made by Johnson also demonstrate the close relationships between the Old World forms and man, the macula lutea tending to disappear as we descend in the scale of New World Monkeys and being absent in the Lemurs. The results which I published upon my tests with precipitins directly supported this evidence, for the reactions