32 
W. E. AGAR. 
rings as formed by splitting of the dyads, but his figures 
show such a close resemblance to the secondary pairing in 
Lepi do siren that it appears to me impossible to escape 
from the conclusion that in Pyrrhocoris the rings are 
formed in the same way, that is, by a pairing of the bipartite 
univalents, and that the change from the monad to the 
dyad appearance during the second synizesis is not due tc 
conjugation, but to the development of a transverse con- 
striction in each separate univalent as we saw so diagram- 
matically in Lepidosiren. Unfortunately Gross does not 
give us the only information which could settle this point. 
He makes no counts of the chromosomes until after the 
stage where ring formation is complete, by which time, of 
course, the reduced number is present. In this respect 
Henking’s account is more complete than Gross's, and seems 
fully to confirm the interpretation which I have ventured to 
suggest. Henking states that in nuclei containing a mixture 
of rings and half rings, the total number of chromosomes, 
counting rings as two and half rings as 6ne, adds up to 
twenty-four (the somatic number). 
Reality of Parasyndesis. 
The debatable stages of the meiotic prophases in which 
parasyndesis and its associated phenomena occur — leptonema, 
zygonema, strepsinema — have been dealt with by many 
experienced cytologists, such as Montgomery, Farmer and 
Moore, Gregoire, A. and K. E. Schreiner, Janssens, Meves, 
Fick, Goldschmidt, Hacker and many others, in whose works 
are to be found full discussions of the evidence for and 
against. 
The establishment of parasyndesis, with subsequent sepa- 
ration of the homologous chromosomes in the first meiotic 
division, may be said to rest on two main foundations. Firstly, 
that the leptotene threads have each the value of a whole 
chromosome, and that they do not represent the temporarily 
separated daughter halves of a precociously split spireme. 
