THE SPERMATOGENESIS OF LEPIDOSIUEX PABADOXA. 31 
less striking than in the case of the long, thin, pre-meiotic 
chromosomes. 
Similar considerations may be applied to all tliose forms 
with simple tetrads divided by a longitudinal and a transverse 
joint, in which the presence of the transverse joint (i. e. the 
one parallel with the spindle axis in the metaphase) is taken 
as evidence of an end-to-end junction of the chromosomes. 
A recent work purporting to describe the origin of the 
tetrads by a simple end-to-end conjugation with subsequent 
longitudinal splitting is that of Gross, who has re-described 
the spermatogenesis of Py rrhocoris. Now an examination 
of his text and figures makes it appear extremely probable 
that the diakinetic stages are really similar to those of 
Lepidos iren. The difficulties which lie in the way of a correct 
interpretation in Py rrhocoris, and the comparative certainty 
with which these stages can be elucidated in Lepidosiren 
is my excuse for attempting to criticise Gross’s account. 
According to this author the first synizesis loosens into long* 
chromosomes, of which the number could not be exactly 
determined, but which were certainly more than the reduced 
number. Now his figures 17 and 18 look very like nuclei 
containing the reduced number of gemini of which the 
members are well separated, though the pairs are still mostly 
distinguishable. As he gives no figures or description of the 
stages before synizesis, which he says follows almost directly 
after the last spermatogonial division, it is impossible to form 
any opinion as to how these gemini have been produced. 
The separation of the gemini (supposing my interpretation 
to be correct) becomes successively more and more complete 
in Gross’ fig's. 19, 20, 21, till they are as well separated as in 
Lepidosiren. After the peculiar temporary disintegration 
of the chromosomes, a second synizesis sets in. As the 
chromosomes enter into this contraction as apparent monads 
and emerge dumb-bell-shaped, Gross assumes that during 
this stage they have conjugated in pairs. Gradually in 
successive stages we find more and more quadripartite rings 
and fewer and fewer dumb-bells. Gross interprets the tetrad 
