1909] GATES--CHROMOSOMES IN OENOTHERA 185 
show the characteristic X, Y, V, H, K shapes, etc. (See figs. 12-14, 
19-21.) 
A large number of counts of the chromosomes were made in these 
telophases, and it was found that the numbers 10 and 11 occurred 
_with approximately equal frequency, while the numbers 9 and 12 were 
only occasional. Every single case which was admitted as a count 
could be determined with absolute certainty as a case in which there 
were, for instance, just 10 chromosomes, no more and no less. 
In the anaphase of the heterotypic mitosis, the chromosomes as 
they pass to the poles are nearly globular or somewhat elongated in 
shape, and are at first closely massed at the poles of the spindle. 
There is considerable variation in the time of appearance of the split in 
these chromosomes, though they usually appear bivalent in the early 
telophase. They appear to come into actual contact at this time, 
forming a compact group, but never fusing or uniting. They very 
soon begin to separate, however, and as they do so nuclear sap appears 
between and to a lesser extent around them. Then the nuclear 
membrane appears where the karyolymph comes in contact with 
the cytoplasm. The nucleus so formed is at first very small, but 
grows rapidly to its full size by the increase in nuclear sap. 
Lawson (20) has described this process in detail in Passiflora 
coerulea. He found that the chromosomes fused into a single mass 
in the telophase, and that the karyolymph begins to be secreted 
within this chromatic mass and later comes to surround it. The 
cytoplasm, coming in contact with the karyolymph, forms the nuclear 
membrane, which is therefore the limiting membrane of the cyto- 
plasm, just as is the tonoplast of a vacuole. It is undoubtedly true 
that the nuclear membrane is formed where the karyolymph comes 
in contact with the cytoplasm, although it must remain uncertain 
whether the karyolymph is secreted by the chromosomes Or merely 
attracted and accumulated about them from the cytoplasm. There 
must be extremely little if any cytoplasm included in the nucleus so 
formed, because the nuclear membrane appears close about the 
chromosome group when the chromosomes are still almost in contact. 
This supports Gricore’s observations (14, 15), that the structural 
contents of the resting nucleus are formed wholly from the chromo- 
somes, and there can be no doubt, in this case at least, that the 
