Igor] GAMETOGENESIS AND FERTILIZATION IN ALBUGO 159 
strictly comparable to that of A. Portulacae and A. Bliti, stage 
for stage. Later the supernumerary oospheric nuclei are elimi- 
nated, leaving only one female pronucleus to function. It seems 
necessary, therefore, to regard the supernumerary oospheric 
nuclei in this species as potential gameto-nuclei, and to regard 
the uninucleate condition in A. Tragopogonis as due to the sup- 
pression of the many and the survival of only one gameto-nucleus. 
In the light of this behavior A. Portulacae must be regarded as 
the primitive form, and A. candida the most highly developed. 
In view of the great functional importance of the coenocentrum 
in A. candida and A. Tragopogonis, it is also rational to regard 
the coenocentrum as an advancing structure. The reverse is 
true, however, in reference to the receptive papilla, as there is 
no indication of its present utility, and everything indicates that 
it is a vestigial structure. The difference in the mode of zona- 
tion between A. Tragopogonis, where the protoplasm aggregates 
into preliminary centers which finally coalesce into one, and A. 
candida, where one central region is formed directly by a reces- 
sion of the protoplasm from the oogonial wall, may well be 
regarded as an outcome of the reduction of A. candida to a 
uninucleate condition of the oosphere, and may be taken as 
further evidence indicative of the highly developed rather than 
the primitive condition of A. candida. It is hardly conceiv- 
able that evolution could have proceeded in the reverse direc- 
tion, that is from the type of oogenesis in A. candida to that of 
A. Tragopogonis, A. Bliti, and A. Portulacae. The serial evi- 
dence from the coenocentrum, the receptive papilla, the mode of 
zonation, and the number of functional nuclei concurs in point- 
ing to the multinucleate conditions shown in A. Portulacae and 
A. litt as primitive, and the uninucleate oosphere of A. candida 
as derived. 
The conditions are almost precisely the same as those in the 
Fucaceae studied by Oltmanns (1889), with the exception that 
in Albuginaceae and Peronosporaceae a coenocentrum plays an 
important réle in the reduction of the number of functional 
nuclei to its ultimate expression. It is not difficult to imagine 
