The spermatogenesis of Peripatus (Peripatopsis) balfouri. 295 



of the dyaster we find the commencement of this union; to what 

 agency is it due? The tension of the mantle fibres would tend to 

 bring into closer contact the central ends than the distal ends of the 

 chromosomes, since were it not for the number and dense arrangement 

 of the chromosomes the tension of the mantle fibres would tend to 

 a strictly radial arrangement of the chromosomes around the pole of 

 the spindle. But no tension of the mantle fibres can explain the close 

 union into pairs which we shall observe in the synapsis stage, namely 

 a union etiected by a short thick band of linin, connecting the central 

 ends of every two chromosomes. Now there is good evidence for 

 concluding that this band of linin is not formed in the synapsis stage 

 for the first time, but that it had persisted continuously from the time 

 of the prophase of the spermatogonic mitosis. 



This evidence is as follows. In the preceding dense spirem stage 

 we found that very probably a single continuous linin spirem thread 

 is present, along which the chromatin becomes evenly arranged. We 

 found, further, that when the chromatin spirem segments into the 

 chromosomes the linin spirem still remains continuous in the form of 

 inter-chromosomal fibres continuous with the linin in the axes of the 

 chromosomes themselves. Further, these inter-chromosomal fibres can 

 still be found in the monaster stage; as in earlier stages these fibres 

 can be perceived only when they lie exactly in the plane of the spindle, 

 so that when they cannot be seen, or only short portions of them be 

 seen, we must consider that they lie in an angle to the surface of 

 the section. We have then even at the monaster stages a persistence 

 of the linin spirem in the form of linin fibres attached to the ends 

 of the chromosomes. In the ensuing metakinesis the chromosomes 

 split longitudinally. In this splitting I conclude that these linin fibres 

 likewise split longitudinally; I must state that I have not seen any 

 stages of such assumed splitting, but I think that any one who follows 

 closely this account of the spermatogenesis will conclude with me 

 that it very probably occurs. It can not seem surprising to assume 

 such a longitudinal splitting of linin fibres, for it is what occurs in 

 every equation division ; in any equation division each chromosome is 

 split longitudinally, but since each daughter chromosome contains linin 

 as well as does the mother chromosome, a longitudinal splitting of 

 the linin within the mother chromosome must take place. Now if 

 that portion of the linin spirem contained within the chromosome can 

 split longitudinally, why cannot that portion which outside of the 

 chromosome composes a inter-chromosomal fibre? In the synapsis 



