" a systematic improvement in the transport facilities of the agricultural areas." For the purpose of carrying out this policy, Sir Eric left the Admiralty when the ministry was reconstructed in Jan. 1919 in order to organize and preside over a new Ministry of Transport. Until the bill constituting it was passed in the summer, he held Cabinet office without portfolio. The measure, which was opposed in both Houses as tending unduly to national- ization, gave the minister control of railways, roads, canals and docks. The constitution of the new ministry subsequently aroused severe criticism on account of its extravagant scale when in 1920-1 all eyes were fixed on the need for economy; and in view of the contemplated handing back of the railways from State control to the companies in the autumn of 1921, it became understood that its future was very doubtful, and that Sir Eric himself had no desire to continue in public life after that event. A bill for regrouping the railways was introduced by him in the spring of 1921. Sir Eric Geddes married in 1900 Gwen- dolen, daughter of the Rev. A. Stokes, and had three sons.
GEIKIE, JAMES (1839-1915), Scottish geologist (see 11.553), died at Edinburgh March i 1915.
GENETICS. This term was proposed at the third conference on hybridization 1906 to denote the study of heredity and variation. In that sense it has been generally adopted, and by extension is understood to include the physiology of reproduction and the art of breeding. Though such inquiries have been pur- sued from the earliest times, the development of a special branch of science relating to them is recent. The primary incentive was the hope that by applying accurate methods of observation and experiment to the course of heredity and variation a more precise knowledge of evolutionary processes might be acquired. Modern theories of evolution are based on the assumption that species have arisen by descent with modification, and that the constancy and diversity which living things manifest in their reproduction provide a sufficient basis for that conception. It is significant that as a result of the preliminary work done under the new inspiration attention has been largely diverted from these more philosophical aims. Beliefs current among naturalists, especially as to the nature and incidence of variability, were at once found to be widely incorrect. The scope and character of these dis- coveries are referred to below. Their immediate consequence has been that the development of evolutionary theory is tacitly suspended or postponed, and activity is concentrated on the exploration of genetical physiology, the theoretical evaluation of the knowledge thus gained being relegated to the future.
In these researches several methods of investigation are available. Modern genetics began with an attempt to observe empirically the course of contemporary variation from type; but though observations of this class have proved valuable in a preliminary survey, and have often been of use as indicating material for more prolonged investigation, the main advances have been accomplished by either (i) experimental breeding or (2) cytology. Important sidelights on genetical problems have also been obtained through the study of developmental mechan- ics (Enlwicklungsmechanik) by experimental methods.
(i) Experimental Breeding. The great stimulus to this method of research was given by the rediscovery in 1900 of Mendel's paper (see 18.115). Heredity, long regarded as a fortuitous and seemingly lawless phenomenon, was proved to follow regular principles which could in great measure be as- certained by experiments properly planned. A vast field was at once thrown open for investigation. Mendel's success was made possible by his genius for simplification. Working with peas he made crosses between distinct varieties and watched the descent of their numerous characteristics, fixing his attention on each separately, and disregarding other differences. He then found that numerous distinctive features behaved in descent as if they were transmitted as units. These determining elements or units are referred to as factors or " genes " (a term especially used by American writers, the equivalent of Johannsen's Genen). The differences determined by these factors can com- monly be shown to be treated in heredity as pairs of alternatives or opposites, such as tall and short, coloured and colourless,
hairy and smooth, each germ-cell being usually pure in respect of one or other of the contrasted characteristics. This is the principle of allelomorphism, and the members of such pairs are called allelomorphs. The zygote, formed by two germ-cells united in fertilization, may be made up of two germ-cells alike in respect of any given pair, in which case it is said to be homozygous in that respect, or it may be a heterozygote if the uniting pair of cells are unlike. Before the germ-cells of the heterozygote are formed a process of segregation occurs, and there is a dissociation between the opposing elements introduced at fertilization, such that the resulting germ-cells are again in normal cases pure in regard to each allelomorph.
After the rediscovery of Mendel's work, progress was rapid, and it was soon found that similar principles of descent apply to a great range of characteristics by which living things are distinguished. The number of forms of life studied is now very large, and includes most of the kinds of plants and animals which are readily amenable to experiment or observation. Man is evidently no exception, and we already know that certain features of human coloration, especially of hair and eyes, and several congenital abnormalities are transmitted according to the Men- delian scheme, some being dominant, others recessive.
Scarcely any satisfactory opportunities for studying the genetics of the lower plants (ferns, mosses, Algae, etc.) have yet occurred, but one example has been described in a unicellular Alga (Pascher). Of the features by which animals and plants are distinguished most have now been shown to be dependent on segregable elements. Reservation must be made in regard to differences which are simply quantitative, for there is a good deal of evidence suggesting that the elements by which size and weight are determined do not often form themselves into simple allelomorphic pairs. A similar doubt exists in regard to numerical or meristic distinctions.
Differences in instinct and other characters dependent on nervous mechanism are not, as such, distinct in their genetical behaviour, and some have been proved to depend on segregable factors or elements. In several breeds of fowls the hens are devoid of maternal instincts, and do not sit on eggs. This charac- teristic is recessive to the normal instinct, and segregation takes place in regard to it. The same is true of the pacing habit in horses as opposed to the trotting habit. The " waltzing " habit of certain Japanese mice is recessive to the normal, segregates from it and breeds true when it reappears. This example is interesting, since the abnormality is almost certainly a conse- quence of deformity in the semicircular canals of the ear.
As to the descent of the normal mental attributes of man little is known with accuracy, but several abnormalities of the nervous system are known to follow modes of descent which prove them to be subject to segregation. Feeble-mindedness is a recessive condition which breeds true. Paralysis agitans is also a recessive. Hereditary chorea descends as a dominant; colour-blindness and a form of night-blindness may also be termed recessive (see SEX). In heterozygous combination with the normal there is segregation, but the descent of these conditions is complicated by sex.
It will readily be understood that though the determining factors may be transmitted as units, the distinguishing charac- ters of animals and plants must be often due to the association of many independent units. Of these some produce their effects separately; but not rarely, though independently transmitted, two or more unit-factors may be complementary to each other and combine to produce a joint effect, or "compound character" as it is sometimes called. Such complementary factors if separate- ly present in the organism without their complement need not manifest their presence at all, and it is then only by breeding tests that their existence can be demonstrated.
Organisms may now be represented as aggregates of units which confer upon them their various attributes. The degree to which an organism may be thus resolved is as yet undetermined, but there is presumably a limit to the process, and it is natural to suppose that the detachable elements are implanted on a basis which for a given type is irreducible.
Reversion. Conceptions, formerly vague, now acquire an
