4i6 
CARL S. HOAR 
for the coniferous origin of the Angiosperms rather than for a deriva- 
tion from the Cycadales. To quote directly from the author's pub- 
lished account: "The idea of cycadalean and bennettitalean affinity 
receives Httle support . from the anatomy of Ephedra. On the other 
hand, there are many points which are opposed to it and in favor of 
the coniferous relationship; the arrangement of the primary vascular 
bundles, double leaf-traces, arrangement and structure of the pits on 
the tracheides, bars of Sanio, tertiary spirals, trabeculae and resin 
plates, primitive uniseriate lignified rays, wood parenchyma, and 
endarch vascular bundles of the leaf. The Gnetales do not appear to 
have arisen from any modern group of Conifers, but rather from or 
close to the base of the coniferous hne." 
"An angiospermous affinity is indicated by the possession of true 
vessels, broad rays, and separation of the leaf-traces upon the stem." 
Figure i shows a tangential view of the broad (aggregate) ray in 
Ephedra californica. 
Among others who have held to the idea of the gymnospermous 
affinity of the Angiosperms we find Eichler (1878) stating that the 
Amentiferae are derived directly from certain gymnospermous families. 
Treub (1891) and others were also of this opinion. 
The next question to arise is which of the families of the Angio- 
sperms are lowest and hence most closely allied to the Gymnosperms. 
Here again a diversity of opinion prevails. Eichler (1890), according 
to the account of Moss (1912), considered Casuarina as belonging 
to the amentiferous forms and as not having peculiarities striking 
enough to warrant a separation from this group. Engler (1897) 
himself so placed the genus until Treub's work appeared in 1891. 
Treub claimed that certain reproductive characteristics made it neces- 
sary to consider the Casuarinaceae as lower than the Amentiferae 
proper. Since that time the family has been placed by Engler's 
system in the Verticillatae at the very bottom of the Dicotyledons. 
Warming (1912) justifies this position in the following words: "By 
the transfusion tissue, by the type of the stomata, and especially the 
ovular peculiarities, the order seems to be related to the Gymnosperms, 
especially the Gnetales." With regard to the ovular peculiarities, he 
found that "in the nucellus, which has two integuments, many embryo- 
sacs are formed, of which only one is fertilized. The egg-apparatus 
consists, as among the Angiosperms, of an egg-cell and two synergidae; 
but antipodal cells are not formed. Before fertilization a prothallus of 
