1911] GATES—CHROMOSOME REDUCTION 337 
However, this is not invariably true, and I have shown that in 
Oenothera Lamarckiana and its mutants, and in O. biennis (4 and 
5, p. 183), the diakinetic chromosomes are usually very loosely 
paired, owing to a weak attraction between homologous chromo- 
somes.’ In O. grandiflora, as I have already pointed out (7), the 
figures of Davis (2) indicate that this attraction is greater than in 
the other forms. 
It is not to be supposed that these suggestions offer a full 
explanation of all the phenomena, but they deserve to be care- 
fully tested in future studies of synapsis. The nuclear enlarge- 
ment unaccompanied by cytoplasmic growth is probably connected 
with the fact that two subsequent mitoses, without further chroma- 
tin growth, follow each other. 
A recent paper by Stomps (21), on reduction in Spinacia, 
offers some particularly accurate figures of the presynaptic stages 
(pl. 1, figs. 7-13), drawings of which have been so frequently more 
or less diagrammatic. Sromps finds that the chromosomes are 
hever joined to form a continuous spirem, but that their free ends 
Can always be determined. There are 12 somatic chromosomes 
in Spinacia, and in the presynaptic nuclear reticulum SToMPs 
believes he is able to determine definitely 6 elongated threadlike 
darker-staining bodies, which are not joined end-to-end, but vari- 
ously arranged. Each of these bodies is considered to be com- 
Posed of two longitudinal portions more or less completely fused, 
Which are interpreted as representing two chromosomes laterally 
Paired. These threads are so delicate. that interpretations are 
extremely difficult, so that the correctness of this interpretation 
must depend upon the accurate demonstration of the number of 
these bodies. In later stages the threads are very much shorter 
and thicker, but the’ figures (pl. 2, figs. 6-13) do not form a close 
fnough series to show whether the final chromosome bivalents are 
formed by a longitudinal or a transverse segmentation of the 6 
(2) bodies represented in the presynaptic stages. If the author's 
"In December 1908, in a paper read before the Botanical Society of America 
(Further studies of Oenotheran cytology; abstract in Science 29:269. 1909), I showed 
t the phenomena of reduction in O. biennis and in O. laevifolia agree in every 
detail with my earlier account (4) of that process in O. Lamarckiana and its mutants. 
