400 EDMUND B. WILSON 



Nevertheless, in these nuclei also the leptotene-threads may often 

 be seen to lie parallel and in pairs on one side of the nucleus, while 

 on the opposite side they are quite irregular. 15 .Here too may be 

 seen T-shaped figures, in some cases almost exactly like those of 

 Batracoseps, except that the stem is more clearly double and shows 

 no indication of torsion. It may again be pointed out that Sut- 

 ton's observations on Brachystola are entirely consistent with a 

 parasynaptic mode of conjugation. I think therefore that the 

 case for telosynapsis in these animals is not yet established, and 

 that in spite of the careful work of Davis, Brunelli and others, 

 the question must still be considered open. 



As to the Hemiptera, sufficient emphasis has already been laid 

 upon the practical difficulties which they present. In Euschistus 

 however, as described in Montgomery's recent valuable paper 

 ('11) the difficulties are less baffling than in many other forms; 

 and he has had the advantage of working with a species in which 

 the total number of threads can be determined in at least some of 

 the nuclei at every stage. In this work the author, reversing his 

 earlier conclusions concerning these insects, definitely accepts the 

 theory of parasynapsis. As I have pointed out, the prophase- 

 figures in Oncopeltus and Protenor are certainly not out of har- 

 mony with this conclusion. I therefore accept the probability 

 of a side by side union in these animals, though I think the possi- 

 bility of an end to end conjugation is not yet excluded. 



But if, now, the fact of a side by side union of parallel lepto- 

 tene-threads be granted, we have still not arrived at a demon- 

 stration of parasynapsis; for there are some very important 

 possibilities yet to be reckoned with. Before such a demon- 

 stration can be admitted, we must first make sure of the number 

 of separate pre-synaptic chromosomes (cf . Fick, Meves) and sec- 

 ondly must exclude the possibility, which has been suggested by 

 several writers, that the parallel union is no more than a reunion 

 of sister-threads that have been derived by an earlier longitudinal 

 fission of a single thread (or chromosome) and have subsequently 



15 This was also noted by Davis but attributed by him to " an accidental arrange- 

 ment, which is more common near the pole since in this region the threads are 

 crowded more closely together" ('08, p. 127). 



