296 Ishikawa.—Studies 011 the Embryo Sac 
the fate of the extra male nucleus in the embryo sac of Oenothera by 
the first alternative proposed above, which was also maintained by Nemec 
in the case of polyspermy in Gagea lutea. 
The second alternative, that is, the production of an extra nucleus 
in the male gametophyte, may be no less applicable than the first one for 
the explanation of the case. Several examples referred to the latter case 
have been reported, one of which is that of Lilinm*aurantiacum examined 
by Chamberlain (7), who found three male nuclei in the pollen grain. 
Another is the production of four male nuclei in the pollen-tube of Ornitho- 
galum and Scilla nutans , studied by Strasburger (65). No such cases have 
been found in Oenothera , but another aberrant case, which will be given 
below, suggests its probable occurrence. Therefore, though not conclusive, 
the writer is inclined to explain the case in question by the first as well as 
the second alternative, until some more conclusive facts are obtained. 
Another aberrant case just mentioned is the fusion of an egg nucleus 
with two male nuclei, while the pole nucleus of the same embryo sac 
receives another male nucleus, as shown in Text-fig. V, 11 , 12 . 11 
shows that the egg nucleus is fusing with two male nuclei on both 
sides, and in 12 two nucleoli of different size are found in a large pole 
nucleus, which is therefore a perfect primary endosperm nucleus. The 
embryo sac from which these figures were drawn was carefully studied, and 
it was found that it had received only one pollen-tube. Thus it cannot be 
denied that an extra male nucleus was produced in the male gametophyte. 
Nemec (49) often observed triple fusion in the egg nucleus in Gagea lutea , 
and proposed several hypotheses for this question, suggesting the origin of 
a plant possessing triploid nuclei as having been derived from an egg nucleus 
fertilized by two male nuclei. Though there is some difficulty in arriving 
at a satisfactory conclusion as to the distribution of the chromosomes at the 
time of the reduction division in the next generation, the origin of the triploid 
mutant must be ascribed to a fusion of diploid and haploid germ nuclei, in 
order to be more rational within the scope of the present knowledge. In this 
respect, Nemec’s view seems to be very ingenious, who substituted two male 
nuclei for the diploid germ nucleus. Some would be opposed to his view, 
advocating that the dispermic egg cannot give rise to a new plant owing to 
the early disintegration. But our present knowledge does not permit us to 
state whether the dispermic egg can develop further or not, and consequently 
we can neither deny nor affirm it. The opponent must be influenced by some 
zoological evidence, for an actual development of the dispermic egg-cell has 
been recorded by several investigators, though some irregularities in the dis¬ 
tribution of the chromosomes in further cell division, may occur, being often 
induced by interference of the extra centrosomes. But in the higher plants 
no such interference does occur even if polyspermy may happen. At any 
rate, Nemec’s view seems to be the most natural one among several 
