34 
W. E. AGAR. 
chromosomes, which separate from each other in meta- 
phase I. 
As was brought out in the descriptive part, the onset of 
strepsinema is obscured in Le pi do siren bv the synchronous 
beginning of synizesis, so that this stage is the least clear of 
any in this animal. Consequently the direct evidence to be 
gained from it, while all in favour of the most usually 
accepted view that the rings are formed by the re-open- 
ing of the longitudinal split temporarily obliterated in the 
pachytene stage, is, it must be admitted, of less value than 
that to be gained from other forms in which this stage is 
clearer. 
Significance of the Temporary Complete Separation of 
THE Ex-CONJUGANTS IN DlAKINESIS. 
The complete separation of the ex-con jugants, which is 
begun in synizesis and completed in late diakinesis, is one of 
the most striking features of the maturation processes in 
Lepidosiren, and a full account has been given of it in the 
descriptive part. Nevertheless, it is only an extreme 
example of the very loose connection between the branches 
of the gemini so often observed in diakinesis. It is only 
necessary to glance through the plates of a few woi'ks on 
spermatogenesis, or still more of oogenesis, to convince one- 
self of this. 
In many forms a diakinetic end-to-end pairing has been 
described, closely similar to the second pairing in Lepidosiren, 
and has been taken for the real conjugation. The case of 
Lepidosiren makes it evident, however, that the mere fact 
that the chromosomes are present in the somatic number in 
diakinesis and then unite end to end in pairs is no proof that 
the parasyndetic scheme does not apply to them. As examples 
of forms with a diakinetic pairing like that of Lepidosiren 
may be mentioned Lumbricus (Calkins), Caloptenus 
(Wilcox), CEnothera (Gates, Davis; in this genus the 
pairing is often a very loose one), and possibly Ophryo- 
