286 
Ishikawa.—Studies on the Embryo Sac 
sacs ; but, as to the twin sacs, only few examples have been given. Coulter 
and Chamberlain ( 13 ) summarized the cases known at that time: two ripe 
sacs in L ilium candidum , Salix, Juglans cordifor mis,Delphinium, and Senecio , 
and two or more sacs in Fagus, Corylus, and Carpinus. Since their publica¬ 
tion several other examples have been reported: mature twin sacs in 
Ulmus americana by Shattuck ( 61 ), in Nelumbo lutea by Cook ( 11 ), in Lychnis 
alba x L. Flos-cuculi by Compton ( 9 ), in Smilacina racemosa by McAllister 
( 41 ), in Peperomia hispidula by Johnson ( 32 ), in Gastrodia elata by Kusano( 36 ), 
and in Godetia Whitney i, Fuchsia ‘ Mar ink a ’, and F. ‘ Emile de Wild email' 
by Tackholm ( 69 ). In addition, the following cases have also been reported : 
three sacs in Cheiranthus Cheiri by Schacht ( 57 ), two in Taraxacum 
officinale by Schwere ( 59 ), the same in Balanophora globosa by Lotsy ( 37 ). 
The writer has also found two sacs in Ludwigia prostrata and Jussieua 
repens , the latter case being illustrated in Text-fig. VIII, 9 a, 9 b, gc. The 
above-mentioned mature double embryo sacs have been explained by the 
previous investigators as having derived from two independent mother-cells, 
except in Nelumbo lutea , in which Cook ( 11 ) stated that ‘ two sacs were found 
one below the other in the main axis of the ovule. The extra sac was 
evidently derived from one of the sister megaspores.’ This is the only 
example of mature twin sacs originated from two sister megaspores, though 
we know many examples in which all or some of the sister megaspores 
develop as far as either a bi- or quadrinucleate condition is reached, but 
only one of the original megaspores becomes a mature embryo sac. 
In Oenothera studied by the writer the twin sacs commonly occur, 
sometimes lying side by side (Text-figs. Ill, 14, 16, and IV), sometimes 
one upon the other (Text-fig. Ill, 13, 15). It seems to be the general rule 
in these plants that of four megaspores the two interjacent ones first 
disintegrate, then one of the survivors follows the example. But, if both 
survivors simultaneously begin to develop, the twin sacs may ultimately be 
attained. This is just the case with Oenothera . No case has been observed, 
however, in which more than two megaspores had developed. 3 and 5, 
Text-fig. Ill, show two successive stages, the former of which represents two 
growing megaspores, and the latter similar ones in a binucleate stage, the 
micropylar one of which is ready for a tetranucleate state. Text-fig. Ill, 6, 
illustrates two developing megaspores in a binucleate condition, though one 
of the nuclei in the lower megaspore is left in the next section. These 
structures generally give rise to twins, as shown in Text-fig. Ill, 13, 15. 
Geerts ( 21 ) states that he once observed an embryo sac which contained 
eight nuclei, and that in the middle portion of the sac a partition line was 
clearly visible. Though he did not give any figure, the sac observed by 
him must have been in the same condition as those described above. 
In Text-fig. Ill, 13, the upper sac, which is already attacked by the 
pollen-tube, is pressing the lower one, while the filiform apparatus is developed 
