94 
Agnes Robertson. 
rather interesting point that the division itself is perfectly sym¬ 
metrical, and it is only the great subsequent growth of the 
functional male nucleus which brings about the inequality. 
Hofmeister (1) gives the number of archegonia in Taxus 
baccata as 5 to 8. Jager (7) confirms this as the normal number, 
but says that he found cases of 9, 10 and 11. In my 1906 material I 
met with an embryo-sac containing 17 archegonia! Two arche¬ 
gonia sometimes occur with no intervening prothallial cells. I have 
not detected any trace of a ventral canal cell nucleus. 
According to Jager, wall-formation does not occur in the pro¬ 
embryo until 16 to 32 nuclei have been formed. My observations 
confirm this. One ovule contained 7 proembryos ! Three of these 
were binucleate, two were 8-nucleate, one shewed 16 daughter 
nuclei still connected in pairs, and in the seventh there were 16 
resting daughter nuclei, but still no trace of wall-formation. 
III.—Affinities and Phylogeny. 
According to the most recent classification of the Taxaceae 
(that of Pilger, 15), this group is divided into three sub-families:— 
1. Podocarpoideae. 
2. Phyllocladoideae. 
3. Taxoideae. 
Other authors however do not place Pliyllocladus in a sub-family of 
its own ; Coulter and Chamberlain (9) include this genus in the 
Taxoideae, while Strasburger (2) attaches it to the Podocarpus 
group. The result of some observations on Pliyllocladus alpinus 
(Robertson, 22) has been to convince me that the older of these 
two views more nearly expresses the true relationship, and that 
Pliyllocladus is intermediate between the Podocarpoideae and 
Taxoideae, hut with somewhat greater affinity for the former. The 
Taxoideae are sub divided by Pilger as follows:— 
Cephalotaxe^:. 
Taxe^e. 
Ceplialotaxus, 
Torreya. 
Taxus. 
Ginkgo used to be included with the genus Ceplialotaxus (cf. Veitch, 
8), but modern work, especially the discovery of the motile sperms, 
has emphasized the uniqueness of the Maidenhair Tree among the 
Conifers, and its resemblance to the Cycads, and it has been 
removed from the near neighbournood of Ceplialotaxus into an 
alliance of its own,—the Ginkgoales. 
