120 



disjunction a very high proportion of 

 3X-superfemales. Present indications, 

 from the uncompleted experiments, are 

 slightly contradictory, but tend to a 

 conclusion which is the opposite of that 

 earlier reported as probable on the 

 basis of the slight evidence then avail- 

 able (Bridges '22). When the number 

 of fourth chromosomes is three the 

 intersexes are more female-like, and 

 u^hen the number is two they are more 

 male-like. The fourth chromosome has 

 a net female tendency, similar to that 

 of the X and different from that of the 

 other autosomes. By variation in the 

 number of fourth chromosomes it is 

 possible to have a fringe of minor sex- 

 types about each of the major types of 

 sex difference. 



The list of sex-types has been en- 

 larged by the discovery of tetraploids, 

 or 4N individuals. These are females, 

 quite identical with normal females in 

 sex characteristics. The tetraploid 

 arose in a stock of triploids; and was 

 detected only by the strikingly differ- 

 ent offspring given. A female supposed 

 to be 3N was selected from the 3N 

 stock and outcrossed to a normal male. 

 All the offspring were triploid females 

 (about 30) or triploid intersexes 

 (about 20). There were no 2N off- 

 spring or supersexes. It was seen that 

 this result might be produced if the 

 mother were 4N instead of 3N. For 

 in that case all the reduced eggs would 

 be 2N; and these fertilized by X-sperm 

 would give 3N females, and fertilized 

 by Y-sperm would give 2X,3A inter- 

 sexes. 



Before the discovery of this 4N in- 

 dividual, an expectation that it would 

 occur had arisen from several facts. 

 Thus, in the three years following the 

 discovery of triploidy there had been 

 found no less than twenty-five in- 

 stances of the new occurrence of tri- 

 ploidy. This very high frequency was 

 paralleled by cytological observations 



BRIDGES 



that give the explanation of the origin 

 of triploids. In three separate prepara- 

 tions of ordinary 2N females it was 

 found that a portion of an ovary was 

 constituted of markedly larger cells; 

 and in two of the individuals some of 

 the giant cells were in division, and the 

 chromosomes could be counted as 4N. 

 Evidently there had been in some 

 oogonial cell a division of the chro- 

 mosomes that had not been followed 

 by division of the nucleus and cyto- 

 plasm. The resulting tissue was tetra- 

 ploid, and any reduced gamete would 

 be 2N. Such a 2N gamete, fertilized 

 by a normal sperm, would account for 

 each of the twenty-five recurrences 

 of triploidy. 



Furthermore, in examining sections 

 of intersexes, two individuals were 

 found in which similar cysts of even 

 larger cells were present. In one of 

 these cysts divisions were occurring; 

 and the chromosomes were clearly 

 6N. A 6N cyst in a 3N female would 

 give, upon reduction, 3N eggs, which, 

 fertilized by X sperm, would give the 

 expected 4N type of female. 



Soon after this first case of 4N fe- 

 male a second similar case was found. 

 Also L. V. Morgan found a third case, 

 and was able to prove by genetic tests 

 that four separate X-chromosomes had 

 been present (in press). 



The fact that 4N individuals are 

 females, not modified as to sex, has 

 important bearings on our ideas as to 

 the way in which genes interact to 

 produce their effect. The view 

 adopted here is that in general the 

 effectiveness is in proportion to num- 

 ber of genes, and the significant point 

 is the ratio between sets of genes that 

 tend to produce alternative effects. On 

 this view we find a ready explanation 

 of the fact that such diverse forms as 

 2N, 3N and 4N individuals are pre- 

 cisely alike in their sexual character- 

 istics; for in all these forms the cffec- 



