545 
Cytology of the A sens, 
attached by one end to the centrosome. On nuclear fusion the centrosomes 
unite, and the chromosomes lie parallel for a time, and in that position fuse 
in pairs. Harper’s studies relate primarily to the fusion in the ascus. 
It is remarkable that in Mildews nuclear fusion, whether sexual or 
asexual, appears to be followed directly by the pairing of the chromosomes ; 
thus the gametophytic number of chromosomes in Phyllactinia (Harper 12)) 
is eight, yet eight appear in the ascogenous hyphae after fertilization 
and eight again in the ascus between the subsequent asexual fusion and 
meiosis. 
Similarly in Humaria granulata union of the chromosomes takes 
place after fertilization ; we have unfortunately no evidence to bring for- 
ward as to the method by which this is accomplished, but the association is 
so intimate that on the spindle of the ascogenous hypha the double nature 
of the pairs cannot be detected. In Phyllactinia a pairing of the already 
bivalent chromosomes to form quadrivalent bodies follows the fusion in the 
ascus ; [in H. granulata this does not occur, but the chromosomes of 
the asexual fusion remain apart till the heterotype division is complete : 
they then pair in the prophases of the homotype. 
Union of the chromosomes brought together by the fusion in the ascus 
takes place at a corresponding stage in Ascobolus furfur aceus and Peziza 
vesiculosa , but in Otidea aurantia it is delayed till the beginning of the third 
division. In Lachnea ster corea, as in Humaria rutilans , no visible pairing 
of the chromosomes occurs in connexion with brachymeiosis. 
The accompanying Text- figure shows diagrammatically the relation 
between the forms described in this paper and some of those previously 
studied. For the sake of uniformity we have represented the minimum 
number of chromosomes as two throughout. 
There is thus a remarkable lack of uniformity in the extent of chromo- 
some union which precedes brachymeiosis. This may be due to the fact 
that the two nuclei which fuse in the ascus are always the products of 
development of a single spore— or at most, if a functional antheridium 
develop on a different mycelium from the ascogonium, of a single pair 
of spores — and they are often much more closely related. Moreover, they 
have been subjected to the same environment, and it appears therefore justi- 
fiable to suppose them identical. Under these conditions interchange of 
material between the chromosomes can have little effect, and the in- 
timacy or duration of their union seems to be without importance. 
Spore-Formation. 
The divergent views held on the method of spore-formation have been 
enumerated in earlier papers (8, 9). In this connexion the chief interest of 
our present investigation lies in the important part played by the vacuolate 
