224 BOTANICAL GAZETTE [SEPTEMBER 
leaf margin; while if the affinities of Podozamites are cycadean at all, 
they are with the Nillsoniales. It seems entirely probable, however, 
that they are coniferous. SEWARD (23) has stated the pros and 
cons of the situation, and in view of the spiral phyllotaxis and bud 
scales at the base of the petiole reaches that conclusion. SCHENK 
(21), on the other hand, compares these scales to those found in the 
living Cycas, and argues that Podozamites cannot be related to the 
conifers as exemplified by Dammara orientalis for three reasons: 
(z) the leaves are not opposite; (2) the vascular tissue in the petiole 
is not like that of Dammara; and (3) the epidermal structure is 
different. The first reason may hold for Zamites distans and 
Dammara orientalis, but it does not hold for Zamites lanceolatus, 
for the original description of this species by LINDLEY and Hutton 
(15) states that the pinnae are ‘‘sometimes opposite and sometimes 
alternate’’; nor does it hold for other species of Dammara, where the 
leaves are spiral. The second reason seems equally questionable, 
for of the two vascular strands figured by SCHENK, one shows pro- 
toxylem rings and the other has the crowded hexagonal pitting 
characteristic of both cycads and araucarians. As regards the third 
reason, the difference in epidermal structure is slight; in both 
Podozamites and Dammara the stomata are in rows between the 
veins, but in the former the long axis is parallel to the edge of the 
leaf, while in the latter it is at right angles. The resemblance to 
Araucaria § CotymBrA, however, seems to be very close. The 
phyllotaxis is the same, and both have rows of stomata with their 
long axes parallel to the leaf margin. Another possibility is pre- 
sented by § NacEta of the genus Podocarpus. It is not suggested 
that Podozamites can be identified specifically with any living 
conifer; for example, the sinuous walls of P. Nageia and A. bra- 
siliana bar them out, as do the heavily pitted epidermal cells of 
A. imbricata and A. Bidwillii; but it does seem fairly clear that 
Podozamites is nearer to the conifers than to the cycads. 
Summary and conclusions 
1. A comparative study of living and fossil conifers indicates 
that epidermal structures are of great value for accurate specific 
diagnoses, but of relatively little importance for indicating affinities. 
