598 
Saxton. — Contributions to the 
used is not entirely satisfactory. It differs considerably from Thomson’s 
(54) figures of the membrane in Cupressoi'deae. The actual thickness of 
the membrane in Tetraclinis is just under 2 \i. 
In a recent paper the writer (47) has discussed the extent to which 
multinucleate prothallus cells are found in the Conifers, in connexion with 
their regular occurrence in the Callitroideae. In Tetraclinis such cells occur 
fairly freely, but frequently show a subsequent fusion of the nuclei within 
the cell. Such a stage was carefully looked for in the genera of Calli- 
troideae, but was never discovered. Figs. 57 and 58 show various unfused, 
fusing, and fused nuclei, all drawn from different parts of the same section. 
There is no difficulty in finding such stages, and it may be taken as certain 
that such a fusion of nuclei would have been readily found if it did occur in 
the Callitroideae. This observation explains Juel’s (21) remarks as to the 
great variation in size of the nuclei of the prothallus cells in Tetraclinis. It 
is not easy to see precisely what is the physiological meaning of these 
nuclear divisions and fusions in the endosperm, but one is tempted to com- 
pare them with the situation in the Welwitschia endosperm (Pearson (40)). 
In detail such a comparison reveals obvious and important differences, but 
yet the same sort of tendency is apparent in both cases, namely, to establish 
the following sequence : Free nuclear division — > nuclear fusion — > forma- 
tion of food material — > absorption of food material by developing embryo. 
The same tendency is seen in the Angiosperms, but whether this justifies 
a morphological comparison between the endosperm in Gymnosperms and 
Angiosperms is another question, into the discussion of which it is not 
proposed to enter. 
Phylogeny. 
Tetraclinis is evidently quite a typical genus of the Cupressoi'deae, but 
nevertheless, there are several points which agree in indicating a high 
degree of probability that it must have been through some plant closely 
resembling Tetraclinis that the southern sub-family Callitroideae were 
derived from the essentially northern sub-family Cupressoi'deae. One of 
the chief regions of distribution of the Cupressoi'deae may be said to be 
Europe and Middle Asia ; Tetraclinis occurs in Northern Africa (Algeria 
and Morocco, chiefly), while one species of Juniper us {J. thuriferci) is also 
found in Algeria, and another species is stated by Drude (13) and Engler (15) 
to occur in tropical Africa. Juniperus appears to be more closely related 
to Tetraclinis than any other genus of Cupressoi'deae, and it is therefore not 
surprising to find species of the former geographically associated with the 
monotypic Tetraclinis . The tropical species of Juniperus (J. procera) and 
W iddringtonia Whytei are both found in the mountains of Central Africa 
at altitudes of from 5,000 to 10,000 feet, while, passing south, we have 
successively W iddringtonia Mahoni , W. Schwarzii , W. juniper oides, and 
