The spermatogenesis of Peripatus (Peripatopsis) balfouri. 297 



deduced quite as mucli from a study of the later stages of spermato- 

 genesis, where the liiiin threads become thicker and more readily 

 distinguishable, as from the preceding stages. As has just been stated, 

 it represents the metakinesis of 4 chromosomes, and the simultaneous 

 splitting of the persisting linin spirem (colored red); the connective 

 fibres {C.F) become simply stretched out without splitting, while all 

 the rest of the linin spirem, that within the chromosomes as well as 

 that connecting different chromosomes, becomes split. We distinguish 

 in this diagram the linin thread joining the central ends of every pair 

 of daughter chromosomes (C.P.L\ as well as the threads which con- 

 nect the distal ends of every two chromosomes (D.P.L). It is by 

 the agency of the former that the central ends of every two daughter 

 chromosomes become united together to produce the bivalent chromo- 

 somes of the synapsis and later stages; the distal linin threads, on 

 the contrary, connect distal ends of separate bivalent chromosomes. 

 What interests us just now is that both these kinds of threads are 

 portions of the original continuing spirem thread, and that they 

 together with the portion of the linin contained within the chromo- 

 somes compose a continuous linin spirem in the daughter cell 

 (1st spermatocyte) in which they lie. Their fate and function will be 

 discussed more fully in the following pages. 



Returning from this digression , it is only in exceptional cases 

 that in the early anaphase the daughter chromosomes become grouped 

 into pairs, i. e. into bivalent chromosomes (Fig. 49 — 52): such cases 

 are precocious. Usually the approximation into pairs is not well 

 marked until the beginning of the synapsis stage, when the nuclear 

 membrane has reappeared, and when the chromosomes have elongated. 



When the daughter chromosomes have reached the poles of the 

 spindle, they are usually so closely crowded together that when deeply 

 stained their individual outlines can scarcely be distinguished (Figs. 57, 

 62); but they never actually fuse together as is proved by studying 

 very thin sections of faintly stained preparations. After this they 

 commence to elongate, their outlines become uneven, and synchronously 

 their structure becomes less compact (Figs. 58, 60, 61, 63). About 

 this time, or shortly afterwards for there seems to be in different 

 cells a slight variation in this regard, the nuclear membrane appears 

 (Fig. 63). Occasionally at this early stage (Figs. 60, 61) each chromo- 

 some shows the commencement of the longitudinal split which later 

 becomes very pronounced; but it is unusual to find this split at this 

 early stage. 



