October 13, 1916] 



SCIENCE 



537 



blindness and some other sex limited affec- 

 tions in man. From an analysis of the de- 

 scents of these characters Morgan concludes 

 that such limitation is in reality only a special 

 case of that complete or partial association of 

 factors in their parental combinations which 

 was first recognized as coupling and repulsion. 

 These phenomena may in fact be all one. 

 They are examples of linkage between factors, 

 the second factor involved in the case of sex- 

 limitation being that for sex. The funda- 

 mental identity of these linkage-phenomena 

 had naturally been suspected. Difficulty, how- 

 ever, lay in the peculiarity of sex-limitation, 

 that in it the linkage has never been observed 

 to be other than complete. The new theory, 

 as will be seen, represents this distinction in 

 a simple and readily conceivable way, so that 

 we are at once attracted. It may be remarked 

 that linkage is no mere incident of technical 

 genetics. We can readily perceive that it must 

 play a great part in the control of heredity. 

 Close resemblances of offspring to parents and 

 grandparents in features and other attributes 

 are common even in families of mixed races 

 like our own. Such resemblances must de- 

 pend on the coexistence of multitudes of fac- 

 tors, and could scarcely ever be perceptible if 

 the factors were really distributed at random 

 among the germ-cells. The theory provides 

 a mechanism by which their associations may 

 be governed. 



From the beginning it was tempting to in- 

 terpret the processes witnessed in the matura- 

 tion of the germ-cells as the visible means by 

 which factors are segregated. Cytologists 

 have shown that when the chromosomes are 

 formed anew from the rested nucleus their 

 number and on the whole their forms are con- 

 stant for the species. They may thus be re- 

 garded as having a permanence or individual- 

 ity. Further, they consist of pairs, one of 

 each pair doubtless representing the material 

 contributed by each parent, the two contribu- 

 tions having retained their identity through 

 all the divisions and changes which have hap- 

 pened since the original fertilization. 



If, therefore, the number of genetic factors 

 were never greater than the gametic or haploid 



number of chromosomes, we should obviously 

 conclude that each chromosome carried one 

 factor, and the ordinary distribution of fac- 

 tors would be produced by a random alloca- 

 tion of one chromosome from each pair to the 

 set comprised in each gamete. But we know 

 that the number of genetic factors in various 

 types of life greatly exceeds the gametic num- 

 ber of chromosomes and consequently this 

 simple account was discarded as insufficient. 

 At this point we meet the first of the far- 

 reaching suggestions which Morgan offers, 

 namely that all the factors are linked together 

 in groups, and that the number of the inde- 

 pendent groups is that of the haploid chromo- 

 somes. This number in Drosophila is four, 

 and it is claimed that, on genetic analysis, 

 the various factors of Drosophila can be 

 proved to be so interrelated as to constitute 

 four linked groups and no more. Before 

 wholly accepting a proposition of such magni- 

 tude we naturally entertain a provisional re- 

 serve, but it may be at once admitted that all 

 the evidence available is capable of this con- 

 struction. Among the animals and plants al- 

 ready studied are many in which the factors, 

 apparently subject to no linkage, in number 

 far exceed that of the haploid chromosomes, 

 but Morgan is able to reply with force that 

 the possibility of linkage in these cases has 

 not been exhaustively investigated. Tests of 

 the heterozygotes by breeding with double re- 

 cessives on a considerable scale provide the 

 only really sufficient method of detecting link- 

 ages. Such work (especially in plants) is 

 commonly very laborious and has rarely been 

 carried out. Thus, though the presumption 

 would a priori seem to be rather against the 

 view that linkage will be found so abundantly 

 operating even in the familiar examples, the 

 speculation is quite legitimate. That it is ex- 

 traordinarily promising as offering at least a 

 chance of positive progress must be obvious 

 to all. 



But if the factors enter the offspring in 

 linked groups — the chromosomes of each pair 

 representing severally the parental combina- 

 tions — the formation of new combinations in- 

 side any one group must mean that there has 



