30 
W. E. AGAR. 
as the division plane in either, and gradually disappearing in 
anaphase II. Now in Lepidosiren it seems certain (apart 
from the correctness or otherwise of my interpretation of the 
previous stages) that the constriction cannot possibly be the 
sign of an end-to-end junction of the chromosomes, seeing that 
it is present in the diakinetic univalents. Probably no one 
will question that these are univalent and not each composed 
of two half chromosomes joined end-to-end. It will be 
remembered that they are widely separated, and in the later 
diakinesis become distributed equi-distantly from one another 
under the nuclear membrane. In this state they may even 
enter into a resting condition. A comparison of the meiotic 
and pre-meiotic nuclei in this condition (PI. 5, figs. 35 and 
36) will probably complete the proof if any be needed. It 
would be difficult to hold that in the one case (fig. 36) 
each chromatin mass has the value of a whole chromosome, 
and in the other (fig. 35) the value of two half chromosomes 
joined together. 
Another difficulty in the way of considering the transverse 
constriction as an indication of an end-to-end junction is its 
almost certain homology with the apices of the Y ? s of the 
spermatogonial or somatic chromosomes. Attention is drawn 
to this on p. 22. In addition to what is said there, it should 
be noted that in the smaller spermatogonia the limbs of the 
V's often seem very loosely connected at the apex (see fig. 7, 
1-6), so that at first glance they look almost like separate 
chromosomes. This reminds us of the apparently nearly 
complete separation of the halves of the tetrads by the trans- 
verse joint so often seen in other forms. , In the larger sperma- 
togonial and somatic divisions the angle between the limbs 
of the V , s is not nearly so sharp — in fact the chromosomes 
often show no tendency to bend at one spot more than 
another. This may account for the unequal lengths of the 
limbs in some of the chromosomes, though, as may be seen 
by reference to the figures, the transverse constrictions do 
sometimes divide the chromosomes into unequal portions even 
in the meiotic divisions, though the inequality is naturally 
