268 THE BOTANICAL MAGAZINE. tV oi; xxiv. No. 2*7. 



abnormal case with a considerable difference was found, and a 

 variable number of chromosomes was also observed in the somatic 

 cell. Besides these poin'.s, some peculiarities in the development 

 of the gametophytes, inclusive the formation of the endosperm 

 in ' Shinriki ' will be described in this paper. 



The material was collected in a rice field in the vicinity of the 

 city of Osaka, and was fixed either with the chrom-acetic mixture 

 or chrom-osmium-acetic mixture. The microtome-sections were 

 made generally 5—10 /j. thicks For staining, Heidenhain's iron- 

 alum-hasmatoxylin was used in general, but sometimes 

 Flemming's safranin-gentian-violet- orange was used. In the 

 former case I have often used congo-red as after-stains for the 

 differential staining of chromatin and linin-substance. 



I. Maiosis in the Development of Pollen-Grains. 



In the young stage of the pollen-motber-cell-nucleus, there 

 are several masses or aggregations of chromatin which Overton 

 (9) has designated as prochromosomes. They are used to lie in 

 pairs (Figs. 1, 2). The number of pairs of these chromatin- 

 masses was found to be nearly equal to the haploid number, 

 though I could not determine it with certainty. In the next 

 stage they begin to grow, take thread-form, and run parallel in 

 pairs (Fig. 3). Then they enter the synapsis stage, in which 

 they conglobe themselves very closely (Fig. 5). In the section 

 cut 2 p. thick, however, the double nature of the threads can 

 be still distinctly observed in this stage (Fig. 6), as the fact 

 itself was also pointed out in other plants by Cardiff (1). 



Whether the conglobation of the nuclear substance at the 

 synapsis is a natural process or not has been a subject of 

 question. Some authors took it for an artifact chiefly caused 

 during the process of fixation, while the others thought 

 it to be a natural phenomenon. Since it was observed 

 by Sargant and other investigators in the living state, how- 

 ever, the matter stands in favour of the latter view ; and at 

 the same time authors generally tend to regard the synapsis 



