September, 1922.] SINOTO—IRIS JAP0N1CA. 101 



as it is in Oenothera (8), Lilium (23) etc, though the nuclear substance 

 of the tapetal cell is occasionally extruded in the narrow space be- 

 tween the pollen mother-cell and the tapetum as seen in Fig. 7. This 

 may probably depend on the absence or disappearance of the plasmo- 

 desms between them. 



There are two views concerning the origin of extruded portions : 

 i. that they originate in chromatin only (Gates, West and 

 Lechmere) ; ii. that they originate in both chromatin and nucleolus 

 (Digby). Kornicke says that the nucleolus does not generally enter 

 the neighbouring cell. In Iris japonica the second alternative appears 

 to be the case (Fig. 5). 



The shape of the extruded bodies is multifarious, and their 

 number on one side of the cell is inconstant. For example, the 

 shapes are globular, pear-shaped, rod-like, crescent, grape-like, irregul- 

 arly massive, etc., and some, as described by Digby and West and 

 Lechmere, protrude secondarily finger-like portions, while others 

 show beaded, granular, or spireme-like appearances, the last being 

 observed by Gates only in Oenothera. In Fig. 3 and Fig. 4 a part 

 of some extruded bodies which is distant from the extrusion pore 

 presents reticulate structures, and one long body penetrating a cell 

 enters the next. The latter case is also shown in Fig. 15 which 

 bears some resembrance to that of the vegetative cell in Iris germanica 

 delineated by Schurhoff(20). Fig. 11 shows a chromatin globule 

 connected with its original mass by a long trailing tail coming into 

 the central region of an adjacent mother-cell. The stage of nuclei in 

 Fig. 6 is one preceding the diakinesis stage, in which the nuclear 

 substance of a cell is not in the centre, and a part of it is extruded 

 in the cytoplasm of the neighbouring cell, and takes an irregular 

 massive form. In Fig. 16 which represents the material obtained from 

 the same loculus as Fig. 6 two fine connections are seen between the 

 extruded body and the main mass. In Fig. 2 a portion of chromatin 

 goes into the cytoplasm progressing in the same direction as that of 

 the contracting nuclear contents. 



The nucleoli in this plant are in general globular even at the 

 time of extrusion of the chromatin (Fig. 1, 2, 3, 5), though various 

 other shapes are also found. Some of the nucleoli in Fig. 3 and Fig. 

 4 seem to be pinned against the cell wall and to take a semi- 



1) In this paper hereafter the word chromatin means any nuclear substance 

 except nucleolus. 



