100 THE ORIGIN OF GYNANDROMORPHS. 



males produced. By prolonging the time to the point where the eggs 

 are almost ready to die, all or almost all of the frogs become males. 

 The result, moreover, appears from Kuschakowitsch's results not to 

 be due to selective mortality. Hertwig attempts to explain the results 

 in accordance with his view of nuclear size versus cell size, but the 

 case seems peculiarly ill suited to this interpretation, because the 

 nucleus has dissolved and the chromosomes are already in the meta- 

 phase condition when the eggs enter the oviduct, and it is here that 

 the delay occurs. It is not at all obvious how delay in this condition 

 can have much to do with cell size versus nuclear size. One of us 

 (Morgan, 1913) has suggested that Hertwig's results may be due to 

 a sort of parthenogenetic development in those eggs whose progress 

 is held back. Such a result might be due either to the egg nucleus 

 giving rise to the embryo (the sperm merely starting it, but taking no 

 further part in the development), or to the sperm nucleus becom- 

 ing the functional one, the egg nucleus having disintegrated in the 

 interval. In support of such a view may be cited the observation 

 of Oscar Hertwig and of Gunther and Paula Hertwig on frogs' eggs 

 treated with radium. They interpret certain of their results as due 

 to mononuclear development of the treated egg or sperm. The sex 

 of the resulting larvae was not determined. The recent results of 

 Loeb and Bancroft and of Loeb have shown that frogs' eggs, in- 

 cited to development by Bataillon's puncture method, give rise to 

 males and females in a few cases in which the frog stage was reached. 

 Loeb states (1918) that the parthenogenetic males have the double 

 number of chromosomes. Herlandt describes the parthenogenetic 

 embryos of the frog as arising in such a way that the haploid number 

 of chromosomes at the first division must be supposed to be present, 

 but Brachet states that he has found the diploid number of chromo- 

 somes present. Until further cytological work is done the explanation 

 of the facts remains obscure. 



Swingle has recently described hermaphroditic stages of the young 

 frog Rana pipiens. Both eggs and sperm are formed in the gonad of 

 some individuals, whereas other individuals have only testes or ovaries, 

 i. €., not mixed. He suggests the possibility that an irregular distri- 

 bution of the sex chromosomes in early oogonial divisions may account 

 for this condition. In one hermaphroditic individual he found 13 chro- 

 mosomes in the spermatocytes, one of which is dumbbell-shaped, and 

 this he thinks is the sex chromosome. In most first spermatocyte 

 divisions the 12 autosomes divide, but the dumbbell-shaped chromo- 

 some goes to one pole. Exceptionally, however, the chromosome 

 divides, one half going to each pole. An irregular division of the kind 

 (or of some other kind), if it occurred at an earlier stage, might give 

 the chromosomal combination that would produce an egg, even in a 

 potential male. 



