DUNN 



causes are to be sought, and, once 

 having arisen by a mysterious occur- 

 rence called mutation, we have learned 

 something of the manner in which the 

 variations are inherited, and by a 

 process of inference have been able to 

 localize still more exactly the region 

 of change. The only permissible gen- 

 erality, then, concerns a general sim- 

 ilarity in the germ plasm and probably 

 in the individual chromosomes of 

 these many species of rodents. But in 

 one case the similarity between species 

 has been found to be more than gen- 

 eral. It has been found to be quite a 

 specific similarit\". 



If we examine this case in detail we 

 find that in two distinct, inter-sterile 

 species, mice and rats, two similarly 

 appearing variations have occurred, al- 

 binism and pink-eye. In rats the genes 

 for these variations are linked with a 

 strength of about 21 per cent, which 

 is possibly slightly in excess of the 

 actual. In mice the linkage between 

 these genes is something less than 15 

 per cent, which is based on observa- 

 tions of 6700 animals raised solely for 

 the purpose of determining this link- 

 age and is probably reasonably accu- 

 rate. In terms of the chromosome 

 hypothesis these facts mean that these 

 two genes are present in the same 

 chromosome in rats and mice, in rats 

 at a distance of 21 units apart, in mice 

 at a distance of about 15 units apart. 

 The difference in location is so small 

 that for practical purposes we can say 

 that they are located at homologous 

 points in the two species. 



In guinea-pigs where both of these 

 variations occur, there is incomplete 

 evidence, but the data which Doctor 

 Wright has supplied indicate that 

 pink-eye and albinism in guinea-pigs 

 are probably not linked and may there- 

 fore be determined in different chro- 

 mosomes. This does not prove, how- 

 ever, that these variations in guinea- 



103 



pigs are not the same as those in mice 

 and rats. It may mean that the chro- 

 mosome which contains both genes in 

 the more primitive Muridae may in 

 the more specialized Caviidae be repre- 

 sented by two chromosomes, the sum 

 of which rather than either one sepa- 

 rately may be homologous with the 

 one chromosome of mice and rats. Al- 

 though this will be recognized as 

 speculation, there is some slight evi- 

 dence that in the evolution of the 

 rodents a fractionation of chromo- 

 somes may have occurred, for the 

 mice and rats have 19 (haploid) while 

 the guinea-pigs have 28. In the rabbit 

 (Oryctolagus cimiculiis) a member 

 now judged too primitive for the true 

 rodents and recently placed in the 

 Lagomorpha with the others of the 

 old rodent suborder Duplicidentata, 

 the chromosome number is probably 

 12. If this progressive increase in the 

 number of chromosomes through the 

 order Rodentia is found to be a fact 

 and not a chance phenomenon asso- 

 ciated with the smallness of the sample 

 of four species from which our cyto- 

 logical evidence is drawn, it may fur- 

 nish a very important clue to a series 

 of evolutionary relationships of more 

 than ordinary interest. 



In the concluding chapter of his 

 recent book Professor Morgan (1919) 

 has referred to the possible evolution- 

 ary significance of the localization of 

 genes as determined by the study of 

 linkage. He has there reviewed some 

 of the work on similar variations in 

 several species of insects by Metz and 

 Sturtevant, pointing out the difficulties 

 to be encountered in applying this 

 method to the analysis of species, chief 

 of which is the necessity of establish- 

 ing the same linear order in each spe- 

 cies of the genes for similar variations. 

 A species in his point of view, and in 

 this he follows De Vries, may ulti- 

 mately prove to be a "community of 



