Gates . — Somatic Mitoses in Oenothera. 1005 
widely separated positions in the equatorial plate. It is, therefore, uncertain 
whether the pairing of chromosomes in the metaphase is but a part of 
a constant association between homologous maternal and paternal chromo- 
somes of the sporophyte, or whether in each mitotic cycle there is a re- 
entering into a paired arrangement of chromosomes which in the resting 
nucleus are less closely or but little associated with each other. The fact 
that in heteromorphic chromosome groups, as in Funkia , Galtonia , Yucca , 
&c., the pairings are always between chromosomes of similar morphology 
is a strong argument for the belief that the paired arrangement represents 
an actual association of homologous maternal and paternal elements. How- 
ever, until it is definitely shown that the chromosomes in gametophytic 
mitoses, such as in pollen grains or in Fern prothallia, have no paired 
arrangement, too much stress cannot be laid upon the general significance 
of the pairing in sporo phytic tissue. Meantime, the evidence from the 
chromosome behaviour in animals, both in spermatogonial divisions and in 
synapsis, very strongly supports the hypothesis of homologous pairing. If 
the gametophytic chromosomes of plants should prove to be also paired in 
metaphase (a view for which I am not aware that any evidence exists 
at the present time), then the explanation would necessarily be a mechanical 
one, and the same explanation would then apply to sporophytic pairings. 
It seems improbable, however, that a purely mechanical explanation of these 
pairings will ever be necessary. 
Another point worth emphasizing is that in the late prophase of 
somatic mitoses a split occurs, just as frequently occurs in the heterotype 
prophase, and may afterwards close up, again as in the heterotype prophase. 
This serves to emphasize a fact which I referred to in a previous paper 
(Gates, Tl a), namely, that the only essential difference between the 
behaviour of the somatic and the heterotypic chromosomes is that the latter 
merely segregate in metaphase, while the former undergo a split. In other 
words, this removes one more of the differences which were formerly 
supposed to exist between the somatic and the heterotype mitoses. 
Again, though the chromosomes usually show a median constriction in 
the telophase, yet this is only of transient duration, and apparently bears no 
relation to the appearance of a split in the succeeding prophase. The 
chromosomes of the telophase pass into the resting condition without any 
indication of a split in their substance, such as has been described for 
certain forms. The earliest appearance of any split in the chromosome 
structure is in the rather late prophase of mitosis. 
Variations in the shape of somatic chromosomes have been compara- 
tively little studied. Nemec (TO, p. 263) has shown that the chromo- 
somes in root tips of Allium montanum , Vicia fab a, and Galtonia candicans 
become much shorter and thicker after treatment with benzine vapour, and 
Kemp (TO) in a careful paper has shown that chloral hydrate produces 
