1 1 8 Salmi . — Certain Archaic Features in the Seed of Taxus 
primitively organized than Torreya, and (2) the Taxales are in their affinities 
nearer to Ginkgo and to the Cordaitales than to any other known plants. 1 
The general similarity which the seeds of Torreya and Cephalotaxus show to 
that of Ginkgo was recognized long ago(Lindley, 1836, p. 317 ; 1853, p. 231) 
by an association of these three genera in the same order (Taxaceae of 
Lindley), till the discovery of motile sperms in Ginkgo led to a separation of 
this genus as the sole survivor of an ancient phylum. The Ginkgoales are 
now universally regarded as being related on the one side to the Cycads, and 
on the other to the Cordaitales ; especially to the latter group, of which 
they may claim to be more or less direct descendants. On the other hand, 
the retention of Torreya and Cephalotaxus in the position long ago assigned 
to them by Lindley (in the order Taxaceae, including such genera as 
Dacrydium and Podocarpus ) has tended to emphasize their affinity with the 
Conifers. I venture to think that this emphasis has led to a false impres- 
sion as to the affinities of the Yews, which, as already suggested, appear to 
lie rather with the Cordaitales and with Ginkgo than with the Podocarps. 
A further discussion of this question with special reference to the position of 
Phyllocladus will be given in my forthcoming paper on Acmopyle, but it may 
be stated here that the view that there is no close relation between the 
Yews and the Podocarps has previously been expressed by several authors, 
notably in a recent article by Dr. Scott (1911). 
Descriptive. 
Ovule . The structure of the young ovule of Taxus is too well known 
to need further description. But the fact is worth recording, that in my 
microtome series of longitudinal sections (in the principal plane) a distinct 
vascular supply comes off from the two main supply bundles, and enters 
the base of the aril (Fig. 1). The aril supply is, however, so inconspicuous 
that it is not surprising that it escaped the notice of the earlier workers 
(Van Tieghem, 1869, p. 281 ; Strasburger, 1872, p. 4), who had only hand- 
sections at their disposal. 
Young Seed . Owing to the hardness of the integument 2 the oldest 
stage of which coherent microtome sections were obtainable is that shown 
in Fig. 2. In this case the aril supply was completely obscured by the mass 
of very dark-staining tissue in that region. In the slightly earlier stage 
shown in Fig. 3 the supply to the aril is seen to consist of a row of scalari- 
form tracheides j phloem could not be identified with certainty, so that the 
orientation of the aril supply is doubtful. In the figure the tracheides are 
shown in solid black, while the dotted region represents thin-walled 
elongated cells, possibly phloem. 
1 See also Lotsy (1911, p. 5 ), where a similar view has been expressed. 
2 It was necessary to keep the sliced seeds in melted paraffin ( 6 o°C. melting-point) for two to 
three months before sectioning. 
