224 AN INTRODUCTION TO MODERN GENETICS 



both of which are recessives; the female is homogametic with a consti- 

 tution ba ha ts^ ts^, while the male is heterogametic ha ha + "- tSo. 

 In the other race a dominant allelomorph Ts^ is used, and in this case 

 the male is homogametic ha ha 4-^^^ + ^'^ and the female hetero- 

 gametic ha ha Ts^ + ^'\ In both cases tassel-seed is the differentiating 

 gene, and we see there is not much difference between male and female 

 heterogamety. We need not be surprised, then, to find that in the 

 family of fish to which Lebistes belongs, in which sex determination 

 also appears to be dependent on a single differential gene, some species 

 like Lebistes itself show male heterogamety, while others, such as 

 Platypoecilius,^ show female heterogamety. 



4. Phenotypic Sex Determination and Hermaphrodites 



Sex determination in most of the organisms we have spoken of is a 

 fairly efficient mechanism; the genetical differences between males and 

 females are sufficient to ensure that intersexes and other aberrant 

 forms only occur rarely. In fish, however, we have seen that the deter- 

 mination is not very stable, so that it is possible to alter its basis by 

 selection. In these circumstances it is not surprising that environmental 

 conditions may become important, and in Xiphophorus, nearly related 

 to Lebistes and Platypoecilius, the genetic sex determination is very 

 weak, or even absent, and is commonly overruled by external agents. 

 The classical example of this type of behaviour, type 3 of our classifi- 

 cation (p. 212), is the Gephyrean worm Bonellia viridis, in which the 

 larvae develop into females if kept isolated, but into males if they 

 become attached to the body of an adult female. ^ The mechanism of the 

 change is still being investigated. There is some evidence^ that male 

 development is provoked by a specific substance secreted by the 

 female, but other authors^ claim that the same effect can be produced 

 by several external agents (e.g. acidity of the sea water, various inorganic 

 salts, lowered oxygen consumption). If we postulate sex realizers at all 

 in this case, we must suppose that they are homozygous in both sexes, 

 or at least that the two allelomorphs in the heterogametic sex are very 

 nearly alike. The variation in degree of intersexuality which can be 

 obtained in larvae treated in the same way (e.g. by a stay of a certain 

 length of time in contact with a female) is sufficient to suggest that 

 there may be some genetic variation in sexual tendencies between 

 different individuals, but there seems to be a considerable range of 



1 Cf. Kosswig 1935. - Rev. Baltzer 1937. 



' Nowinski 1934. * Herbst 1935. 



