Review. 
204 
origin and development of the embryo-sac and of the pollen mother¬ 
cell which is recorded during the past seven years. 
The authors show very good grounds for the belief—in which 
they follow Overton—that the mother-cell of the embryo-sac is 
morphologically a megaspore, and that where a single megaspore 
occurs it is the survivor of a primitive group of four (p. 75). Such 
groups are shown to occur in many Angiosperms, and they are in 
all probability homologous with the almost universal tetrad of 
microspores. 
In dealing with structures so far reduced as the Angiospermous 
gametophyte, it is often difficult to define their limits with exactness. 
Strasburger first pointed out in 1894 that the reduced number of 
chromosomes in the karyokmetic division of their nuclei was 
characteristic—so far as was then known—of the gametophyte 
throughout the vegetable kingdom. Later work has confirmed this 
statement, though, owing to difficulties of observation, the bulk of 
evidence is still very small, and it follows logically that the male 
gametophyte of Angiosperms dates from the pollen mother-cell, and 
the female gametophyte from the mother-cell of the embryo-sac 
(p. 41). The male gametophyte is much reduced, but there can be 
no doubt that its development ends with the formation of the two 
generative nuclei. 
It is not equally clear what structures represent the female 
gametophyte of Angiosperms. For the present the question is well 
described by the authors as a “ morphological puzzle ” (p. 3). 
Nothing could be better than the way in which this difficult problem 
is handled in Chaps. V., VII., and VIII. The main issue is, of 
course, the morphological interpretation of the endosperm, and/ the 
alternative views are presented here with perfect fairness, while the 
authors make no attempt to conceal their own preferences. One 
criticism is suggested by this discussion. In dealing with the 
question of triple fusion the authors lay some stress on instances in 
which the polar nuclei never meet, but divide independently to form 
endosperm nuclei. On reference to a previous paragraph (pp. 166-7) 
five cases are found and they are quoted without any qualification 
which might suggest that the embryo-sacs were not normal in other 
respects. But on looking up the references the reader finds that 
the structure of the embryo-sac is more or less abnormal in each of 
the five cases. In Autennaria alpina fertilization does not take 
place: the egg developes pathenogenetically. In Balanophova no 
fertilization occurs: the embryo is developed apogamously from the 
endosperm. Helosis in all probability resembles its near ally 
