I 



278 BOTANICAL GAZETTE [October 



later passes out into the egg cytoplasm. Sometimes this fragmenta- 

 tion is accompanied by kinoplasmic fibers, as shown in fig. 10. Fig. g 

 also shows an intact jacket-cell nucleus which has passed into the egg 

 cytoplasm. Occasionallyj though rarely, instances were noted in 

 which the jacket cells do not become binucleate and break down at 

 the time of fertilization; some of them, especially those near the 

 middle of the archegonium, have the nucleus surrounded by delicate 

 cytoplasmic radiations similar to those noted in the unfertilized egg. 

 These nuclei also very much resemble the egg except in size, being 

 of course much smaller. Apogamy was at first suspected, but further 



i 



study showed that these egglike jacket nuclei do not function, being 

 finally broken down and absorbed by the embryos. 



The second male nucleus is normally in the position shown in 

 fig. 5. It enlarges and occupies the upper part of the egg and extends 

 into the space recently filled by the adjacent jacket cells. The chro- 

 matin is arranged in strands, at the ends of some of which it may 

 be aggregated in small groups {figs. 6, 7). The nuclear membrane 

 breaks down on one side {fig, 6) and cleavage planes pass up into the 

 nucleus. Fig. 7 shows a male nucleus with an irregular mass of 

 spindles with groups of chromatin granules at the poles. Later the 

 region occupied by the male nucleus is filled with a number of small 

 cells which may have come from further division of the second male 

 nucleus, or some of them may be a joint product of chromatin from 

 the male nucleus and from the jacket nuclei. Below this region are 

 numerous small cells which possibly may have resulted either from 

 further division of the functionless procmbryos, or by division of the 

 jacket nuclei which have wandered into the egg cytoplasm or from 

 both sources, and still others from masses of cytoplasm which may 

 have been separated into smaller masses of cytoplasm by cleavage 

 planes, as shown in fig. p. Numbers of the very small cells seem to 

 have a single granule in the center; others appear to be only masses 

 of cytoplasm. This nutritive mass, presumably derived from so many 

 sources, is very ephemeral and is quickly absorbed by the functioning 

 procmbryos. 



In the ephemeral nutritive mass in which the procmbryos lie it 

 seems that we have at least a suggestion of the origin of endosperm. 



i 



f 



i 



The term endosperm is here used in the strict angiosperm sense and 



