86 PROCEEDINGS: BOSTON SOCIETY NATURAL HISTORY. 
be seen in the same held. Presumably the causes of elongation in 
this case are the same as in the preceding ones. There is no evi¬ 
dence to show that some of the astral rays become attached to the 
plasma membrane and then contract, thereby drawing the nucleus 
up against the membrane. 
Another striking anaphase condition is the persistence of the 
asters after the spindle fibers have disappeared. But this happens 
with the last generation of daughter nuclei only. They persist until 
the young spores are formed. This naturally leads us to enquire 
what part, if any, the astral rays play in spore delimitation. 
This question has been investigated by Harper (’ 97 ) in a number 
of forms of which jErysiphe communis was apparently best suited 
for the purpose. His answer in a word was that the astral rays 
effect the separation of the spore plasma from the surrounding pro¬ 
toplasm by forming the limiting membrane or plasma membrane of 
the young spore. He states that the eight nuclei of the third gen¬ 
eration lie irregularly arranged in the asciis, each with its centrosome 
and polar rays. A beak then appears on each nucleus on the end of 
which is the aster. This beak is about equal in length to the diam¬ 
eter of the nucleus and is a small tube sometimes with an expanded 
top. Into this tube the chromatin network extends. His assump¬ 
tion was that this last fact seemed to prove that there was a union 
between the chromatin and the centrosome. He regarded the beak 
as a growth and not as being produced by a pulling out of a side of 
the nuclear wall. He noted that the beaks were turned toward the 
ascus wall. As soon as ( before , according to his ngures) the beaks 
had reached their maximum length, the polar rays experienced a 
remarkable transformation. The ra 3 T s began to turn outward, and 
in moving outward came into contact laterally and fused into a 
membrane. At this time the membrane formed a funnel or inverted 
hollow cone the apex of which was at the centrosome. The rays 
constituting the membrane now began to grow rapidly in length 
and at the same time to bend down toward the nucleus. The mem¬ 
brane cut through the protoplasm and finally its edges met at a 
point opposite the centrosome. This completed the process of 
delimitation and at the same time furnished the spore with a plasma 
membrane. 
Turning now to Hydnobolites we find that the phenomena pre¬ 
sented there are apparently quite different. We have already 
