STUDIES ON CHROMOSOMES 397 



case one may indeed often be in doubt, particularly in the earlier 

 stages, though many perfectly clear F-figures are evident. Such 

 doubtful cases may however very well be due either to a confu- 

 sion produced by threads of linin, to defects of fixation, or to 

 coagulation-products of the nuclear enchylema. Batracoseps 

 seems to me, however, decidedly more favorable for study of this 

 point than Tomopteris, partly because of the much greater size 

 of the nuclei, partly because of the greater brilliancy of the pic- 

 tures in detail, especially evident in material fixed with Carnoy's 

 fluid. At its best this method, in my experience, is much supe- 

 rior to Flemming's or Hermann's fluids for study of this point. 

 Prolonged search among the hug.e amphitene-nuclei of Batra- 

 coseps has failed to show even a single clear case in which more 

 than two leptotene-threads can be traced into connection with a 

 single pachytene. When the latter terminate anti-polewards in 

 more than one leptotene-thread two are always seen, very often 

 diverging like the branches of a F; and these bifid figures appear 

 with the utmost clearness in every view — sidewise, endwise and 

 obliquely. It is of course true that such bifid figures are often 

 not in evidence. Not infrequently pachytene-threads seem to 

 end abruptly without connection with the leptotene; sometimes 

 they seem to end in single leptotene-threads. But in the nature 

 of the case the true relation of the latter to the former must often 

 fail to appear in the sections. This may result from many causes 

 -7-accidents of sectioning, entanglement of the threads, unfavor- 

 able position, and the like — and it is very probable that in the 

 coagulation-process of fixation the delicate thin threads may often 

 break away from their normal connections. When allowance is 

 made for these sources of error it is in fact surprising that so many 

 demonstrative F-figures are seen; and it is a significant fact that 

 these figures, though often bent or distorted, always show the 

 same orientation in the nucleus with respect to the centrosome 

 pole. 



That the F-figures represent the normal and typical relation of 

 the pachytene-threads to the leptotene seems to me indisputable ; 

 and I consider it utterly impossible to interpret these figures as 

 an expression of a progressive longitudinal splitting of previously 



