July, 1922] 
CLELAND OENOTHERA FRANCISCANA 
are fully formed by the reverse process of condensation before a split 
occurs. I have come strongly to the conclusion that the spireme in the 
heterotypic prophase of this plant is formed by an irregular process of 
condensation, rather than by the approximation of two distinct threads. 
This conclusion not only conflicts with the idea of an early spht in the 
chromosomes of this plant, preparatory to the homoeotypic mitosis, but 
also with that of the possibility of synapsis, or side-by-side approximation 
of whole chromosomes, taking place during this period. 
During synizesis, the process can be seen to be a continuation of that 
described above. Little indication of paralleHsm of any kind can be found, 
and what little bit there is, is only what one might expect from chance, 
or from the unequal condensation of the various threads of the system. 
When the spireme is finally formed, all the evidence seems to support 
the theory that it is univalent in nature. From late synizesis to the end 
of the second contraction stage, it remains uniformly and evenly single. 
As it begins to condense and shorten, so that the long, tangled loops into 
which it is thrown stand out more and more clearly, the single nature 
becomes all the more noticeable, for it is at this period that the split, in 
many plants and animals, becomes especially evident, and the separated 
threads twist around one another forming the characteristic "strepsinema " 
condition. In this plant, however, there is no true strepsinema, or twisting 
of threads, for there is no split in the spireme. The split in preparation 
for the homoeotypic mitosis does not appear in this plant, except rarely, 
until the middle or end of interkinesis. 
As the spireme becomes more and more massed in the center of the 
nucleus during second contraction, the loops become very short and thick, 
and it can be seen that the chromosomes forming the loops are attached 
end to end. These loops can be traced throughout the whole of the second 
contraction period, and it seems quite certain that the ring-shaped bivalent 
chromosomes which emerge from the knot are the loops which entered 
it. Each bivalent seems to have come, therefore, from an unsplit section 
of the spireme. It is true that toward the end of the second contraction 
period I have found a very few cases in which there seemed to be some 
evidence of a split. These instances were so rare, however, in comparison 
with the cases in which the thread seemed perfectly whole, that the only 
interpretation possible is that they represent a premature split (premature 
for this plant, though usual in most if the telosynaptic view is correct) 
in preparation for the homoeotypic division. 
Diakinesis also presents strong evidence for the telosynaptic arrange- 
ment of chromosomes in the constant and uniform linking together of 
bivalent chromosomes, a phenomenon which can hardly be explained on 
a parasynaptic basis. The most reasonable explanation seems to be that 
the bivalent chromosomes represent sections of the spireme which occupied 
such a position in the nucleus that, when the whole system became con- 
