PHENOMENA ACCOMPANYING FERTILIZATION 319 



In Paramecium, therefore, the first two divisions of the miero- 

 nuclei in endomixis correspond to the reducing divisions in conjuga- 

 tion, the third division as absent in aurelia but present in caudatum. 

 Ivanic (192S) described a similar nuclear history during the processes 

 of encystment of Chilodon uncinatus. If reduction occurs with the 

 first two divisions the four products in endomixis are equivalent to 

 haploid nuclei so far as the chromosomes are concerned, and corre- 

 spond, therefore, to the first type of parthenogenesis above. But 

 they are likewise equivalent to the fertilization nucleus and develop 

 with the diploid number of chromosomes. This number unfor- 

 tunately, is too large in Paramecium to permit of accurate counting, 

 while in ciliates with a small number of chromosomes, endomixis 

 takes place during encystment where cytological details have not 

 been made out in any case. Fermor (1912) indeed described the 

 union of the two macronuclei and of the two micronuclei in Stylo- 

 nychia pustulata during encystment, but the account of the phe- 

 nomenon is incomplete and on its face implies the fusion of diploid 

 nuclei. This is so improbable from the chromosome standpoint 

 that the result cannot be accepted without confirmation. Later 

 work by Ilowaisky (192(i) failed entirely to confirm Fermor's inter- 

 pretation of the happenings during encystment of Stylonychia. 



As indicated above (p. 303) the difficulty over haploid and diploid 

 chromosome number reaches an extreme in connection with the 

 third division of the ciliate nucleus. If reduction in number occurs 

 during the first two meiotic divisions then the pronuclei are formed 

 by a third division of an haploid number of chromosomes. If 

 this division is transverse as appears to be the case with Para- 

 mecium, this third division might also be a reducing division, and 

 the amphinucleus coming from the union of such nuclei would 

 be haploid. If the third division, however, is equational the pro- 

 nuclei would still have the haploid number and their fusion would 

 result in a diploid amphinucleus. The latter appears to be the 

 correct solution. Gregory (1923) for example describes 24 dumb- 

 bell-shaped chromosomes in the nuclear plate of the first meiotic 

 division of Oxytricha fallax. This number is reduced to 12 dumb- 

 bell-shaped chromosomes with this first division and each dumb- 

 bell divides longitudinally. The equational halves are separated at 

 the second division and 12 dumb-bells form the equatorial plate 

 of the third division (Fig. 1()2). The two halves of the dumb-bell 

 are finally separated with this third division, 12 single chromosomes 

 passing to each pole. The pronuclei thus have 12 single chromo- 

 somes and the amphinucleus formed by their union has 24. An 

 equivalent process occurs in Uroleptus halseyi where there are 48 

 chromosomes in the first division reduced to 24. These 24 are 

 reduced to 12 in the second division and these 12 are divided trans- 

 versely in the third division (Figs. 151 , 153) . The interpretation here 



