396 EDMUND B. WILSON 



1. That the 'amphitene' stage (to employ Janssens's appro- 

 priate term for the synaptic nuclei) is preceded by one in which 

 the nucleus is traversed by fine, undivided, leptotene threads the 

 free ends of which are from an early period polarized towards the 

 pole of the nucleus near which lie the centrioles. 



2. That from this pole, during the amphitene stage, thick and 

 often plainly double threads are formed progressively towards the 

 opposite pole near which (in Batracoseps) lies the chromoplast. 



3. That pari passu with the growth of the thick threads the 

 thin threads disappear until all have vanished. 



The conclusion is irresistible, and will hardly be disputed, that 

 the thick (pachytene) threads grow at the expense of material 

 supplied by the thin ones (leptotene). 



It is further indisputable that in many cases the thick (and 

 often double) threads terminate anti-polewards in two undivided 

 diverging thin threads like the branches of a F, which often sepa- 

 rate at a wide angle and may be traced for a long distance, some- 

 times to opposite sides of the nucleus, as continuous threads. 

 This fact may be seen in both Tomopteris and Batracoseps with a 

 clearness that admits of no doubt. These F-figures are so numer- 

 ous, so clear, and in their more striking forms so different from 

 anything seen at other stages as to constitute a highly character- 

 istic feature of the nuclei at this particular stage. 



Janssens, the Schreiners, Gr4goire and others have with good 

 reason insisted on the fact, seen with especial clearness in Batra- 

 coseps, that not more than two thin threads are thus seen diverging 

 from the anti-poleward ends of the thick threads. Pick ('07) after 

 examination of the Schreiners' preparations of Tomopteris, stated 

 that he could sometimes observe more than two such diverging 

 threads. Even Janssens in his earlier work (with Dumez) on 

 Plethodon, believed that he had seen a similar appearance "Un 

 chromosome naissant est parfois en relation avec plusieurs fila- 

 ments, de tel maniere qu-il devient tres difficile a 1'observateur 

 de faire un choix" ('03, p. 423) ; but in his later work on Batra- 

 coseps he insists that such is not the case. 



I have studied this point with the greatest care of which I am 

 capable in both Tomopteris and Batracoseps. In the former 



