W. T. Saxton. 
250 
As regards the megaspore membrane , only one investigator 
(Thomson) has made a comparative study of that structure through¬ 
out the Conifers; his results chiefly emphasized the fact that the 
Araucarians are not very nearly related to any other Conifers. 
It will thus be seen that an analysis of a number of characters 
concerned with embedded reproductive structures reveals the fact 
that only three of these (structure of male gametophyte, position of 
archegonia in prothallus, and structure of the pro-embryo) have a 
sufficient number of distinctive types to be used advantageously in 
the classification of the group. 
The relative value of various anatomical characters has not 
been discussed here; anatomical characters as a whole are held 
(for reasons stated above) to be of inferior importance to characters 
concerned with embedded parts of reproductive structures, and 
moreover it appears as though there were but little agreement 
amongst the various students of Conifer anatomy as to the bearing 
of the known facts on questions of phylogeny and classification. It 
will however be readily conceded that such important points as the 
multiseriate pitting of Araucarian metaxylem tracheids, and the 
combined pitting and spiral thickening in Taxad metaxylem, cannot 
be overlooked in any scheme of classification; there is even less 
danger that the claims of external characters of the reproductive 
organs to be used for classificatory purposes will not be recognised. 
Consequently, in using the characters which have been discussed 
as the sole basis of classification, there is no intention to entirely 
disregard anatomical and external reproductive characters; on the 
contrary, it will be found that these confirm, as a rule, the evidence 
furnished by gametophytes and pro-embryo. 
III.— Various Schemes of Classification of Conifers. 
As long ago as 1872 Strasburger published a scheme of 
classification of Conifers in which Sequoia and Sciadopitys were 
made the types of two tribes (the second monotypic) ranking equally 
with the Abieteas. In the third of his series of Gymnosperm studies 
in 1900, Arnoldi (as far as one can judge, independently) again 
proposed to separate Sequoia and Sciadopitys as the types of two 
independent tribes, both monotypic in his scheme. It is perhaps 
curious that both these proposals were made before the facts which 
completely justify such a suggestion were fully known, but it cannot 
be too strongly urged, in the opinion of the writer, that there can 
no longer be any doubt whatever of the correctness of the view that 
both these genera must be separated in this way. On this view 
