ipoS] YAMANOUCHI—SPOROGENESIS IN NEFHRODIUM 19 



tion of cliromosomes and that the chromosomes of the first 



an 



of the loops; while the second division is longitudinal. Atkinson 

 (4) found two different methods in Arisaema and in Trillium; in 

 the former species the first division was transverse, as Dixon and 



* ScHAFFNER f ound ; 

 SxRASBtTRGER and 



transverse 



pcctivcl}' 



there occur two longitudinal divisions, that the first division is 



is ion 



transverse, the third was 



tion. The authors (Strasburger and Mottier) who described 

 the existence of such a case changed their opinion and returned to 

 their former view that both divisions are longitudinal, so that 



persist only two \' 



mm 



claiming that the first is transverse 



bv GUIGN.\RD (27), GrEGOIRE (22) 



Lloyd (31), and others. 



However, the more recent studies of synapsis have mvolved ver>' 

 detailed and accurate observations concerning the nature of the reduc- 

 tion division. Some of the investigators mentioned abo^•e have 

 changed their view^s, while others still maintain them, but in both 

 cases a new or modified notion concerning: the origin of chromosomes 

 and the nature of reduction division is h 



"•''' v^^'o 



As described before, the nucleus of Xephrodium in the presynaptic 

 stage consists of a complex, anastomosing ragged chromatin reticulum. 

 This ragged reticulum shows a tendency to become transformed into 

 a thread structure, but the process does not occur simultaneously in 

 different regions. Whm the transformation has taken place, the 

 two parts of the thread are observed running side by side from the 

 first. Such a condition as is represented in figs. 16 and i6a is evidently 

 what was called "leptonema" by Winiwarter (51). Further trans- 

 formation of the thread structure from the ragged reticulum results 

 in a nucleus with a continuous chromatin thread or spirem whose 

 double nature is only i-isible at certain parts on account of the clt^ 

 association. The thread becomes tangled and contracted in one 

 side of the nuclear cavity {figs. 17 and 17a) until finally there results 



