THE FACTORS WHICH DETERMINE SEX 671 



either to arrenotokous females or to more thelytokous females, and 

 the proportion of arrenotokous females so produced is liable to con- 

 siderable variation. Maupas ^ has sought to connect this variation 

 with differences in temperature, and Nussbaum ^ with differences in 

 nutrition, but neither conclusion has been satisfactorily established. 

 The question has been reinvestigated by Punnett,^ who has carried 

 out a number of further experiments. In one of these a strain 

 which had hitherto appeared to be purely thelytokous was subjected 

 to considerable fluctuations of temperature. The rate of reproduction 

 was much retarded, but in the subsequent generations which were 

 produced no arrenotokous females could be found. Starvation 

 experiments were undertaken, and in these also thelytokous females 

 which had hitherto " bred true " continued to do so. Punnett 

 concludes that neither temperature nor nutrition has any influence 

 in determining the production of arrenotokous females. On the 

 contrary, it is the property of certain females to produce arrenotokous 

 females in a definite ratio, and the property of others to produce none. 



Theories which assiome that the Gametes are themselves Sexual. — 

 Many biologists have entertained the conception that the gametes 

 are themselves sexual, and a number of facts have been adduced 

 which give very strong support to this idea. Some of these have 

 already been mentioned, but probably the strongest evidence in favour 

 of this generalisation is that relating to the existence of special 

 sex chrbmosomes contained in the ova or spermatozoa. 



Seller* has shown in Talaiporia tubulosa that the direction in 

 which the maturation spindle turns during division determines 

 whether the sex chromosome remains in the egg or passes out in the 

 polar body .8 If it remains in, then the egg is male determined, and 

 if it passes out, female. The interest of his observations lies in the 

 evidence he has adduced that this condition can be experimentally 

 controlled by heat or cold. If the eggs are kept in the cold, 

 3° to 5° C, a higher proportion of females appear (ratio 155 ? to 100 S), 



' Maupas, loc. cit. See also Lenssen, loc. cit. 



^ Nussbaum, "Die Entstehung des Geschlechts bei Hydatina seiita," Arch, 

 f. Mikr. Anat, vol. xlix., 1897. 



^ Punnett, " Sex-determination in Hydatina," Froo. Roy. Soc, B., vol. Ixxviii., 

 1906. 



* Seller (J.), " Das Verhalten der G-eschlechtschromosomen bei Lepidopteren," 

 Arch. f. Zellforsah., vol. xiii., 1914. "Geschleohtschromosomenuntersuchungen 

 an Psychiden," Zeitschr.f. ind. Abst. u. Vererhungdehre, vol. xviii., 1917. 



^ In this respect Morgan's observation on Phylloxera {Jour, of Exp. Zool., 

 vol. vii., 1909) shotild be recalled. It would seem that here the behaviour 

 of the sex chromosomes is controlled by some factor, for it can be seen in 

 certain obgonial divisions that one daughter cell (on account of its small size) 

 has already had its sex determined before the chromosomes have commenced 

 to draw apart and move into their respective daughter cells, so that the 

 chronlosome complex of the daughter cells can hardly be claimed as the sex- 

 determining factor in this instance, but more properly as a sex resultant. 



