96 BOTANICAL GAZETTE | AUGUST 
has been a growth of new fibers. It is not impossible that some 
of these extend into the central spindle, thus increasing the 
appearance of zonal differentiation, but I could not establish 
this by actual observation. In an earlier paper Guignard 
figured such a relation but did not describe it. It seems tobe 
not improbable that such a condition may exist, and that such 
fibers may grow in length and form new connecting fibers, 
either by fibers from the opposite nuclei fusing, or by a com 
tinuous growth from one nucleus to the other. This process 
would account for the apparent increase in number of connect 
ing fibers. Still the evidence is too meager to lead to aly 
definite conclusion. There is no convincing evidence that there 
is any real increase in the number of connecting fibers. Its 
appearance may be due entirely to the changes which take 
place in the original fibers, as is the case in the larch. On the 
other hand, the relatively small number of the connecting fibers _ 
in earlier stages seems to show that some new ones have beet 
formed. The question needs further investigation. If there 
are new connecting fibers formed they seem to act simultaneously — 
with the original fibers in the process of forming the cell plate 
The point that seems of most importance to me here is the pre ! 
viously described relation of the radiating fibers to the daughter 
nuclei, 2. ¢., that they center not upon the poles of the spindle 
but upon the nuclei themselves. This relation, combined with 
the fact that the fibers are new formations, may indicate that | 
the nuclei are the metabolic centers for the formation of spi : f 
fibers. Such an hypothesis is further strengthened by the 
viously described changes in the connecting fibers in the laf 
where the increased thickness is first evident near the ends 
the fibers, 7. ¢., in those portions nearest the nuclei. The * 
that the appearance of the new fibers comes prior to the recom 
struction of the daughter nuclei does not invalidate the ab0Y 
hypothesis, for, as Juel has shown in the formation of abort 
pollen grains in Hemerocallis fulva, single chromosomes | 
have spindles formed between them and the normal nucle” 
Here would seem to be a case in which a chromosome 4S su : 
