(4) Variation by Loss and by Addition. Whether the evidences of authentic variation remaining after the deduction of spurious testimony has been made suffice as a basis of evolutionary theory has been questioned by competent naturalists. Lotsy, for ex- ample, maintains that we have no proof of contemporary varia- tion arising otherwise than as a consequence of crossing; and apart from such extreme pronouncements it is noticeable that as regards varieties of animals and plants anciently domesticated, modern authorities usually incline to ascribe a multiple origin even for forms like wheat, the fowls, pigeons, sheep, horses, etc., which used formerly to pass for derivatives from single types, a belief which is now felt to be inconsistent with what is known of the limits of variability. Distinction must be made between recessive and dominant variations, arising respectively by loss and by addition. As regards recessive variations arising by loss of elements few will doubt the adequacy of the records (e.g. in the sweet pea, Primula sinensis, etc.). As regards the de novo appearance of dominant characters the evidence is less abundant. Morgan and the American geneticists have made prominent several instances of this kind in Drosophila (fruit fly), of which the spontaneous origin of " eosin " (a new and peculiar pink) eyes in a white-eyed strain may be cited. Admitting provision- ally these examples as free from objection they are nevertheless extraordinary events and not common occurrences. Were the dominant in question one already familiar we should hesitate to believe in its spontaneous origin. That a pea genetically vvTinkled, having the characteristic starch of the wrinkled varieties, should without crossing produce a variety with " round " starch-grains would in modern lights appear not much less im- probable than the spontaneous generation of life.
But, as explained in the article GENETICS, nothing absolutely forbids us from inverting the representation of positive and negative factors by extension of the conception of inhibitors of which many are familiarly known; so that we may express the apparent addition of a new element as a loss of one which when present had repressed the new attribute. This symbolism, though admittedly objectionable when dominance is complete, does without strain apply to all cases in which the heterozygote is intermediate, and a large range of alleged new dominants can be covered. In so far as this conception applies, evolution is con- ceived of as a process of unpacking, a progress consisting in the loss of component elements.
(5) Mutation. The term mutation introduced by de Vries is now generally accepted to denote definite genetical variations which are sensibly discontinuous. Though contemporary ex- amples which satisfy all tests are not abundant, there is no question that they occur and have occurred in most of the forms of life. They are indeed part of the occasional experience of most breeders of animals and plants. The special example, the Oenotheras, on which de Vries mainly founded his own theory was singularly unfortunate, and must, as explained in the article GENETICS, be now discarded as inapplicable. Mendclian analysis was only lately made known and the group of discoveries com- prised in the term genetics were in an incipient stage; nor had the criteria of genetic purity, which must be applied to a parental form before the production of new types from it can be accepted as proof of original mutation, been clearly established. The Oenotheras which produce the presumed mutations are now proved to be no pure genotypes, and the suggestion that they were in a " period of mutation " arose from a misunderstanding of the nature and consequences of heterozygosis.
(6) Inter-specific Sterility. The new forms whose productions we witness are never new species. In Primula sinensis about 20 pairs of factorial differences have been determined, which in their several combinations present an amazing polymorphism. A systematist, if he met these forms in nature, might and probably would quite justifiably take many of them for distinct species. But interbred, they and their products are perfectly fertile. Polymorphism like this is, even in a state of nature, far more abundant and far greater than the evolutionists of the last century imagined, yet it avails us little as material out of which true specific differences can be supposed to develop. The con-
spicuous defect in the evidence for the origin of species by com- mon descent remains. Though much is known as to the incidence of variation, not rarely of a magnitude which might naturally be claimed as constituting specific difference, no one has yet raised types from a common origin which when interbred pro- duce sterility of the kind and degree which is one of the common- est attributes of crosses between natural species. By whatever concatenation of arguments theories of evolution have been constructed, that most essential link has never been supplied. The lapse of time is occasionally invoked in the hope of rectifying this and similar evidential defects, a strain which has been main- tained distinct for a long period being thought more likely to show interracial sterility when crossed with its progenitor than one newly separated. Reasoning of this kind, plausible enough in scholastic days, is not acceptable in an age of chemistry, nor may we suppose that that which is never begun will be attained by mere effluxion of time. The more genetical experience extends, the more serious does this hiatus in the evidence become.
(7) Evolution. In allusion to this and other difficulties, which genetic research has forced into prominence, the question is sometimes asked whether the theory of evolution holds its place so firmly as it did, or, more crudely put, whether Mendelian discovery has not "upset Darwinism." It should therefore be stated explicitly that in spite of all the objections with which the doctrine of the origin of species by descent is now seen to be beset objections of which the strength is far more clearly known than before and though as to the manner by which new species have come into existence geneticists adopt for the most part an agnostic attitude, yet all agree that the lines of argument converging to support the theory of common origin are so forcible and so many that no alternative can be entertained. The geolog- ical record is conclusive. To take one most cogent instance: if Angiosperms had existed in the carboniferous age their remains must have been preserved; therefore Angiosperms have arisen since that time, and we cannot conceive whence they came if not by descent from the preexisting plants. Common descent, though rarely if ever a proposition demonstrable in any detail, ranks as an axiom. For Darwin and any other evolutionist before or after him this is a concession of the main claim. Parts of the apparatus by which the validity of that claim was enhanced have fallen into desuetude. In particular the modern geneticist assigns to natural selection a subordinate and inconsiderable role. Or- ganisms are to be considered as coordinated systems. That each particular structure or instinct comprised in the system, which shows permanence or definiteness, makes a contribution to the success of the system equivalent to the cost of its production is_ recognized as a fallacy. We are also reluctant to apply to the interrelations of the collective properties of organisms argu- ments which would be out of place in similar considerations of the attributes of unorganized substances. We no more look for utility in the details of a peacock's feather than in the iridescence of a Roman bottle or in the regularity of basaltic prisms.
(8) Adaptation. It is not merely in regard to the mode by which species have arisen that agnosticism has prevailed. While unwilling to accept adaptation, with Darwin, as a summation of happy accidents, we have no alternative to offer, nor is there in the recent attempts of various experimenters to find that organ- isms transmit to their posterity structural emendations in re- sponse to parental experience anything which sensibly alleviates the difficulty. Most of these claims are obviously faulty and few require serious notice.
Each step in the progress of this branch of science has rather compelled the recognition of genetic determinism; and the hope that by change in the conditions of life or by any external in- fluences significant alteration can be induced in succeeding generations, whether of organisms amenable to experiment or of the human population, must be abandoned.
(9) Classification. The full implications of factorial analysis in relation to biological classification are not yet appreciated. The fundamental idea of the systematist, that animals and plants can be grouped into species, and that the distinctions between species are of a different order from those characteristic of van-
