Cardiff: Synapsis and reduction 277 



finds constant. Montgomery explains this union of whole chromo- 

 somes as a pairing of maternal with paternal chromosomes. These 

 bivalents enter the contraction stage, become longitudinally split, 

 • and in the first mitosis are divided transversely, thus separating 

 whole somatic chromosomes, while in the second mitosis the 

 division is longitudinal, an equation division. Montgomery ('03, 

 '05) confirmed these observations in later papers. 



Sutton ('02) reports an end to end fusion in Brachystola, an 

 orthopteran, in the late telophase in the last spermatogonial 

 mitosis, but finds that the separation of these chromosomes does 

 not take place until the second division following their union ; 

 thus the first mitosis is equational, the second reductional. 



Dublin ('05**) in Pedicellina, a bryozoan, describes an end to 

 end fusion at the end of spermatogonial and oogonial divisions 

 and a separation of these, reduction, in the first mitosis. The 

 second mitosis is equational. Haecker ('04) reported results 

 essentially similar to those of Dublin ('o5<?) and Montgomery ('01). 



Gregory ('04) in several Pteridopliyta reports a formation of 

 bivalent V-shaped chromosomes in the reduced number. These 

 segment transversely with the first division and longitudinally in 

 the second division, thus in essential agreement with Schafther 

 (97) and Farmer and Moore ('05). Moore in later paper (Moore 

 and Walker, '06) reports results in agreement with his earlier work. 



Another group of recent investigators — Allen, Berghs, Greg- 

 oire, Miyake, Overton, Rosenberg, Schoenfeld, Schreiner and 

 Schreiner, Miss Stevens, Winiwarter — obtain results somewhat 

 at variance with those previously mentioned. Allen ('04), in a 

 preliminary note on the microspore formation in IMium, finds that 

 after the formation of the mother-cell there is a long growth 

 period in which the chromatin is in a reticulate condition. Dur- 

 ing the lattet; part of the growth period the chromatin changes 

 from irregular reticulate masses into threads which become ar- 

 ranged in pairs, the moieties of each pair parallel, as they pass 

 into synapsis. In synapsis the chromatin is massed in a tight 

 knot at one side of the nucleus, often pressed against the nuclear 

 wall with the nucleolus. As these parallel threads enter synapsis 

 they move closer together and finally fuse to form a single thread 

 which shows no evidence of its bivalent character for some time 



