226 Nichols, A morphological study of Juniperus communis var. depressa. 



nuclear thread appears as a slehder, deeply staining band of uni- 

 form thickness (flg. 82). The spindle is at first multipolar, but it 

 soon becomes a multipolar diarch, the upper extremity of which is 

 very broad, while the lower bundles of fibers converge toward the 

 center of the asteroid, thus giving to the whole figure somewhat 

 the form of an inverted, truncated cone (fig. 83). With the seg- 

 mentation of the spirem and the orientation of the chromosomes at 

 the equator the nuclear membrane disappears, leaving the spindle 

 surrounded by the cytoplasm of the central cell, and the multipolar 

 diarch soon becomes bipolar. The axis of the spindle may be 

 parallel to that of the archegonium, but more frequently it is in- 

 clined at an angle. Frequently the spindle is some distance below 

 the neck cells. 



The daughter chromosomes, as they approach the poles, are 

 U or V shaped (figs. 84, 85). At the poles the chromosomes 

 rapidly draw together, becoming separated from the surrounding 

 cytoplasm by membranes, and two resting nuclei are developed. 

 No cell plate is formed, nor is there any indication whatever that 

 a wall is ever developed between the ventral canal nucleus and 

 the egg nucleus. 



The two nuclei are at first very similar in appearance, but 

 the egg -nucleus matures rapidly, while the ventral canal nucleus 

 develops slowly and as a rule disintegrates before fertilization. 

 Usually the ventral canal nucleus lies above the egg nucleus, but 

 one case was found (flg. 91) in which these relations were reversed, 

 a condition previously described by Coker (1903 b) in Taxodium. 

 Figs. 86 and 87 show two unusually well developed ventral canal 

 nuclei. No evidence was found to support the theory (cf. Cham- 

 berlainl899) that this nucleus is the homologue of the egg nucleus, 

 but very often, as noted by Coker (1902, 1903b) in Podocarpus 

 and Taxodium, by Land (1902) in Thuja, and by Noren (1907) 

 in J. communis, the ventral canal nucleus persists for a long time. 

 Figs. 88 and 89 show such nuclei in archegonia where the egg 

 has been fertilized and the development of the proembryo has 

 begun. Sometimes they undergo division (fig. 90) and this division 

 is mitotic. 



Incidentally it may be noted that in this species archegonia 

 occasionally are found superposed one above the other (fig. 91) or 

 more or less deeply imbedded in the prothallium. Similar cases 

 are described in Picea and Äbies (Miyake 1903a, 1903b), and in 

 Pinus (Ferguson 1904), while analogous conditions occur in the 

 Bryophytes (Coker 1903a). 



Surroundiug the archegonium complex is a fairly well defined 

 layer of jacket cells (fig. 95), while the walls of the egg cells are 

 thin, except in the upper part, and no structures are visible which 

 can be interpreted as pits. Stop es and Fujii (1906), however, 

 have recently shown that even in those Gymnosperms where the 

 thick wall of the egg is perforated by pits, the latter are closed 

 by definite membranes which are pierced only by delicate threads 

 of cytoplasm, so that the actual transfer of solid substances 



