colour-complex. Until systematic crossing was undertaken, the extremes existed but the intergrades did not. So also in Drosophila, of which the normal eye is red, a profusion of intergrades ranging to the white eye, which was discovered first, has now appeared. Though "mutation" is involved, the essential change is probably the dis- integration or fractionation of the originally integral complex.
(2) Cytological Interpretations of Genetic Phenomena. Soon after the rediscovery of Mendelian analysis the plausible suggestion was made that the behaviour of the chromosome in the course of the maturation divisions was consistent with what might be expected if they were actually the bearers of segregable factors. Since, however, the number of segregating factors in many forms far exceeds the number of chromosomes possessed by those forms, it is clear that if the chromosomes are the carriers of factors they must be capable of carrying many. The discovery of linkage, and especially of the fact that linkage was determined by the parental associations of the factors, pointed in the same direction, for, as hinted (by Punnett) in the earlier article on Mendelism (see 18.118), linkage or "gametic coupling," as it was then called, might not unreasonably be supposed to be based on . chromosomal association. The first development of this conception was made by T. H. Morgan, whose investigations, relating mainly to the fruit-fly Drosophila, have inaugurated a new phase in the development of genetical theory. This .insect is a subject unusually favourable for experiment inasmuch as it offers a profusion of variations or " mutations," and reproduces itself with great rapidity under laboratory conditions.
The work began with the observation that the eyes, normally red, may be white, and that this variation is sex-linked, behaving genetically precisely as colour-blindness does in man. The white- eyed male mated with normal females produces offspring all normal. Of these the sons cannot transmit the abnormality at all, whereas the daughters mated with normal males transmit the white eye to half their sons. White-eyed females can only be produced as daugh- ters of white-eyed fathers and all the sons of such females are white- eyed. Supposing the male to possess an X-chromosome, this system of descent would be represented if it were assumed that in the nor- mal the X-chromosome carried the dominant factor for red eye (see SEX). The linkage with sex is thus found to be an expression of the association of the two determining factors for sex and red eye in the same chromosome.
Numerous other sex-linked characters were soon after discov- ered, to which the same considerations apply, all collectively compos- ing one linkage-group. The other factors identified in Drosophila, amounting to more than a hundred, can all be represented as grouped in three separate linkage-systems which, with the sex-linked group, make four; and since from cytological observations the haploid number of chromosomes in this animal is also 4, the inference is drawn that the factors composing each linkage-group are borne in one chromosome. Developing this conception, Morgan suggests that the factors are arranged in the chromosomes as beads on a string, each having a position normally fixed in relation to the rest. Cross- ing-over is thus represented as the consequence of an exchange of material between homologous pairs of chromosomes in synapsis (see CYTOLOGY). The pairs of chromosomes which then conjugate are with much probability regarded as respectively of maternal and paternal origin. The conjugating pairs seem to twist round each other, and occasionally there is (especially in Amphibia) an appear- ance of anastomosis between them which is regarded as providing for an exchange of material between the homologous pairs, and thus for the formation of cross-overs. According as the linkage between two factors is more or less complete it is supposed that the distance between the position of the two factors in the chromosome is smaller or greater, and in proportion as factors are placed close together the probability of their being separated in the process of twisting and anastomosis is regarded as diminished. The proportion of cross- overs is thus taken as a measure of the position of two factors in the chromosome. If A, B, and Care three factors in one linkage group, and the closeness of the linkages between A and B and between B and C respectively be determined experimentally, then from these two the linkage between A and C can be calculated, and the result of the calculation is commonly found to agree with the value found experimentally for that linkage. In this way the relative " loci " of - numerous factors have been determined with fair consistency, and the fact that this can be done forms a strong argument for the belief that in some way at least the factors must be disposed in linear sys- tems. That these systems are actually arranged along the lines of the chromosomes is as yet a matter of inference. Attention must be called to the curious fact that in Drosophila crossing-over never occurs in the males in any of the 4 linkage-systems. As in every example of sex-linkage studied, the linkage with the sex-factor is always complete; but all the other factors are liable to crossing- over in the female, though among the male gametes the original parental .combinations reappear unchanged. Conversely Ta,naka,
examining linkages in the silkworm, observed that a pair of linked factors show crossing-over in the male but not in the female, and the two facts together suggest some limitation of crossing-over to the sex which is homozygous in sex, the female in Diptera, the male in Lepidoptera. The development of the idea here outlined has become the subject of very active research and is described in a copious but somewhat esoteric literature which can be followed only with difficul- ty by those not personally engaged in the work. That the outcome of these researches has led to a valuable codification of genetic prin- ciples is not in dispute; but until the main thesis, that the number of independent factors or of linkage-systems is never greater than the haploid number of chromosomes, has been shown to hold gen- erally for animals and plants, this account of the nature of linkage, though probable, cannot be regarded as proved. The defect of the theory at the present time is that it rests on many subordinate hy- potheses which are hot all capable of independent verification. The position of the factors, for example, is believed to be liable to changes due to the action of other factors, the effects of age and miscellaneous influences difficult to distinguish. Errors of cell-division, long regard- ed as the most probable source of variation, may also cause disturb- ance. In two very remarkable instances it has been found possible to connect a disturbance in the normal course of heredity with a visible cytological irregularity called by Bridges "non-disjunction." In a certain family he observed that a sex-linked character failed to follow its normal distribution to the sexes, and he was able to find that in this family the sex-chromosomes showed corresponding irregularities. More recently (1921) he obtained similar evidence in regard to the fourth chromosome and the group of genes attributed to it. Thus a definite association between particular chromosomes and the transferable factors must certainly exist.
Giant-forms. The interrelation of genetical and cytological phenomena is further illustrated by the behaviour of " Giant- forms." This name is applied to certain varieties (chiefly of plants) in which the haploid and diploid numbers of chromosomes are double those of normal forms. R. P. Gregory bred such vari- eties of Primula sinensis, and found that in respect of various allelomorphs they might be quadripartite and not merely bi- partite as the normals are. A plant, for example, might be DRRR in colour or leaf-shape, and, in consequence of the extra recessive elements, not distinguishable from the ordinary re- cessive, though in fact capable of throwing a small proportion of dominants. Since recent cytological studies have shown that series of allied forms may contain various multiples of the low- est haploid number (Chrysanthemum, for instance, having 9, 18, 27, 36 or 45), various extensions on these lines may be expected.
Somatic Segregation. In the genetics of plants a number of phenomena have been encountered which are difficult to recon- cile with the view, otherwise not unacceptable, that the distribu- tion of the factors occurs exclusively in the maturation processes of the germ-cells. Apart from certain special conditions, best known in variegated plants (which are sometimes irregular mosaics and sometimes consist of an outer " skin " and an inner " core," dissimilar in their genetical potentialities), there are many plants in which the distribution of factors must have been laid down before the formation of germ-cells. E. R. Saunders's results proved that in certain stocks (Malthiola) the pollen all carried doubleness though the ovules were mixed in character, single and double. C. Pellew showed that in the hermaphrodite Campanula carpatica " pehiformis " the pollen bore exclusively femaleness and preponderantly white flower-colour (the plant being heterozygous for blue). The pollen of Begonia Davisii (a wild species with single flowers) carries doubleness exclusively, and several similar examples are known, in all of which the segregation of characters must precede the maturation of the germ-cells. Thus, while it is not in question that segregation depends on some cell-division, and very possibly on a differentia- tion of the chromosomes, there is evidence that the cell-division in which this differentiation occurs must at least sometimes pre- cede germ-formation. As mentioned, in Oenothera this " unilat- eral " distribution is exceptionally frequent.
Bearing on Evolutionary Theory. This aspect of genetics can only be briefly treated here (see also under MENDELISM). Genetic analysis has shown that the appearance of variability as a con- temporary and widespread phenomenon is largely illusory. On studying a variable species critically it is found that the various forms cannot all produce each other as was formerly assumed, but that they stand in a regular descending order,
