50 THE BOTANICAL MAGAZINE. [Voi. xxii. No. 254. 



tube enters the archegonium and fuses with the egg-nucleus. 

 The fusing nuclei are surrounded by a dense zone of starch- 

 granules (Fig. 12). The fusion-nucleus soon divides (Fig. 13). 

 The two segmentation-nuclei now move down towards the base 

 of the archegonium. 



The further process of proembryo-formation does not seem 

 to differ much from what has deen observed in Taxodiuni and 

 Cryptomeria. The w^all-formation between the free nuclei seems 

 to take place right after the third division, at the bottom of 

 the archegonium. The proembryo consists of three tiers of cells 

 and the middle tier elongates into the suspensor while the upper 

 tier, which is open above, forms a rosette, and the lovsrer tier 

 develops into the embryo (Fig. 14') , , 



The present stud}^ shows that the gametophytes and em- 

 bryogeny of Cunninghatnia shows a close affinity with Taxodiuni 

 and Cryptomeria and are distinctly of the Cupresseae type. 

 And so far as the embryological evidence goes, I can only 

 confirm the suggestion of Arnoldi'^ followed by Lawson and 

 and CoKER that, among the family TaxodiccC, Cryptomeria, 

 Taxodium and Cunninghamia should be placed with Cupresseee, 

 and that Sequoia and Sciadopytis should each constitute a 

 family by itself. 



Botanical Institute, Agricultural College, 

 Imperial University, Tokyo. 



1) Beitriige zur Morphologie einiger Gymno^permen : V. Weitere Untersuchungen 

 der Embryogenie in der P'arailie der Sequoiaceen. Bull. Soc. Imp. Nat. Moscoii. 1901. 

 Rei^rint p. 22. 



