J. W. H. Harrison 273 



genes. On the contrary, in the autumnata $ x dilutata cT Fi lot no 

 pairing of chromosomes takes place at all, and in consequence all of the 

 maturation divisions are equational. Every gamete then derived from 

 the autumnata $ x dilutata cT males is endowed with the sum of the 

 haploid chromosome numbers of autumnata and dilutata, i.e. it has 

 30 + 38 (= 68). In other words it has a full complement of those from 

 each parent. When such gametes encounter in fertilisation those of 

 dilutata a zygote is generated the cells of which contain a chromosome 

 number of 30 + 30 + 38 (= 98) — a double set of dilutata and a single set 

 of autumnata chromosomes. Now if the chromosomes are the heredity 

 bearers and the various unit factors are located in them then the genes 

 from dilutata are duplicated. But Nilsson-Ehle^ and Shull^ have shown 

 in their experiments on chaff colour in grain and capsule shape in Gap- 

 sella that duplicate genes are not necessarily cumulative in effect ; 

 whence it follows that in back crosses such as these, where no genuine 

 reduction division takes place in the F^ insects, the phenotype of the 

 back cross may not vary perceptibly from that of the Fi insect. Further- 

 more, since in the gametogenesis of the back cross pairing of homologous 

 chromosomes will take place, its gametes will be identically the same 

 as those of the Fi insect so that reversal to parental type can never be 

 brought about in the case of hybrids with such a form of gametogenesis. 

 No matter through how many generations we pursue the experiments 

 the constitution of the genotype produced can never differ from that 

 obtained in the first back cross ; thus a constant intermediate form will 

 be maintained. 



Of the larvae only five pupated, bacterial disease carrying off the 

 rest. All of these were males and this is precisely the condition of the 

 dilutata $ x autumnata ^ broods in general ; still, in view of the small 

 numbers involved no definite importance can be attached to the fact. 

 Before emergence three died simply through lack of viability, and two 

 began to develop but were sacrificed for cytological purposes. However, 

 this back cross has once more been obtained and I hope to pursue my 

 investigation with it during the present season. 



1 Nilsson-Ehle, " Einige Ergebnisse von Kreuzunger bei Hafer und Weizen," Bot. 

 Notiser (1908). Nilsson-Ehle, " Kreuzungsuntersuchungen an Hafer und Weizen," Lunds 

 Univ. Irsskr. N. S. Vol. v. pp. 1—122 (1908). 



* ShuU, "Duplicate Genes for Capsule Form in Capsella barsa-pastorU," Zeitsch. fiir 

 Abst. und Vererb. Vol. xii. pp. 97 — 149 (1914). 



