1909] GRIGGS—MITOSIS IN SYNCHYTRIUM 351 
stances would be something of a novelty, but it would not necessarily 
involve functions of the chromatin differing fundamentally from some 
with which we are more familiar. Whenever the chromatin is with- 
drawn granule by granule from a karyosome (nucleolus) preparatory 
to mitosis, we must suppose, on the individuality hypothesis, that the 
granules come out of the nucleus as they went in, properly sorted, so 
that each granule is taken into that chromosome from which it came. 
On the assumption of an equational division in heteroschizis and 
nuclear gemmation, we should have to assume, in addition to this 
well-known property of the chromatin granules to sort themselves out - 
from the apparently homogeneous mass of chromatin of the karyo- 
some, only the further property of reproducing themselves while yet 
within the karyosome as they ordinarily do in the spirem. This is 
perhaps not too much to ascribe to the chromatin, but even such an 
addition to our theory would add a very large field for speculation 
where the conclusions, with our present methods, could never be 
checked by observation. 
But when nuclear gemmation takes place in spirem two alternative 
Possibilities are presented, the choice between which depends on the 
nature of the spirem involved. If the spirem is an entirely different 
Structure from the mitotic spirem, it would be rational to suppose 
that each small karyosome, i. e., each granule of the spirem in nuclei 
Such as fig. 37, contains four chromosomes derived by an equational 
division of the mother karyosome as in the previous cases. But if the 
amitotic spirem is like the mitotic, the situation is entirely different, 
for we know from its later history the composition of the mitotic 
spirem. It contains altogether only four chromosomes, which are 
arranged serially, and any small part will contain chromatin from 
only one chromosome, if a small part of such a spirem is extruded 
and forms an independent nucleus. Therefore, it is evident that 
ts chromatin is derived not from all four but from only one chromo- 
Some. It is then a matter of great importance to determine the con- 
Stitution of the spirem, but unfortunately, as stated above, the evidence 
on this point is somewhat ambiguous, and a definite decision in favor 
of either alternative seems impossible. 
Certain features of nuclei dividing by constriction, however, seem 
'0 throw some light on this matter. When amitosis by constriction 
