402 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1914. 
every aspect of a real sexual process, as De Bary had asserted in 
1863. What made Harper’s discovery still more significant was the 
determination of the fate of the fusion nucleus in relation to the 
nuclei of the ascus and spores. Harper found that one of the row of 
five or six cells resulting from the division of the fertilized o6gonium 
has two nuclei. These two descendants of the diploid nucleus, 
formed at fertilization, afterwards fuse, and the cell containing them 
swells to form the single ascus of this species. This, presumably 
tetraploid, fusion nucleus of the ascus then grows and divides three 
times to give the eight spore nuclei. In the following year or two 
Harper (1896-97) demonstrated a sexual fusion of the same type at 
the initiation of the fruits of another mildew Erysibe, and of the 
saucer fungus Ascobolus. The numerous asci of these forms all, 
arise from binucleate branches of the binucleate, subterminal cell of 
the fertilized o6gonium. Each ascus is at first binucleate, but later, 
as had been seen by Dangeard (1894), the two fuse and then by 
division the eight spore nuclei are formed as in Spheerotheca. 
In the course of the following decade Harper reported the occur- 
rence of two nuclear fusions, like those of Spherotheca, and at the 
same points in the life cycle, in the mildews Erysibe (1896) and 
Phyllactinia (1905), and in the saucer fungus Pyronema (1900). 
Moreover, he found in Phyllactinia a synapsis and evidences of a 
double reduction of the chromosome number in the divisions of the 
presumably tetraploid, fusion nucleus of the ascus. Pyronema 
proved interesting also in having multinucleate gametes, such as 
were at this time being studied by Stevens in the white rust Albugo. 
Harper believed that many pairs of male and female nuclei fuse in 
the odgonium of Pyronema. 
As the outcome of this whole series of studies by Harper it seemed 
clear that there is in many Ascomycetes an alternation of a haploid 
generation, the vegetative mycelium and the sterile hyphe of the 
fruit, with a diploid generation, the fertilized o6gonium and the 
ascus-forming hyphe arising from it. The second fusion, in the 
ascus, was regarded as a nutritive phenomenon to provide a nucleus 
adequate in size for the organization of the relatively huge ascus. 
At the opening of the century the observations of a number of 
- workers on the simpler Ascomycetes, e. g., those of Juel (1902) and 
Barker (1902), seemed to establish the occurrence of a nuclear fusion 
in the odgonium in these forms also. This, with Harper’s work, 
made it seem probable that this fusion is a frequent phenomenon 
throughout all the Ascomycetes. 
The researches of certain other cytological workers, however, 
convinced them that no fusion of nuclei really occurs in the oé6gonium 
of the Ascomycetes which they studied. Thus, Dangeard (1897 and 
1907), working on Spherotheca and Erysibe, found no fusion except 
