the Ovule of My rica Gale. 361 
supports the latter view. The ovules of the orders Julianiaceae, Myricaceae, 
and Juglandaceae — especially of the two latter— are very similar, retaining, 
if the foregoing is correct, ancestral characters. The Julianiaceae show one 
step in advance of the other two orders in the change from the orthotropous 
to hemi-anatropous type of ovule, thus obliterating one of the ‘ ancestral 
characters ’ described in the other two orders. The Amentiferae show 
a further advance in this respect, the ovule being truly anatropous. As 
a result of this change in the position of the ovule, probably more 
pressure came to bear on the parts of the ovule than they would be 
subjected to in the orthotropous condition, the integument and nucellus 
therefore came into closer contact, and since the integument was not large, 
and did not become succulent, there was no longer any need for an integu- 
mental vascular system. 
Although the Amentiferae have advanced further than the Myricaceae 
and Juglandaceae in respect to the structure of the ovule, they show a more 
primitive state of affairs than do the latter in the development of the ovule, 
examples of chalazogamy, numerous macrospores, &c., being found as 
described above. 
Thus in the structure and development of the ovule these closely 
allied orders show, as is shown in the evolution of all groups of plants, an 
unequal development of parts. The Amentiferae which have advanced 
farthest in the specialization of the ovule, losing such ancestral characters as 
the free nucellus and integumentary vascular supply, still retain a primitive 
mode of development of the embryo-sac ; while others such as the 
Myricaceae which show these primitive characters in the structure of 
the ovule, have advanced to the normal Angiospermous type of develop- 
ment of the embryo-sac. 
The evidence derived from the structure and development of the ovule 
in the series of plants including Juglandaceae, Myricaceae, Amentiferae, 
Casuarinaceae, and Julianiaceae, seems to me to support the view that they 
are a primitive group of Angiosperms, and should therefore retain the 
position assigned to them by Engler in his scheme of classification. 
The alternative suggestion made by Arber and Parkin 1 and by Hallier 2 
that the Amentiferae are degenerate orders descended from ancestors of 
the Ranales type is not supported by the facts obtained in my investigation 
of Myrica Gale , for the development of vascular supply in the integument 
can hardly be regarded as a step in the degeneration of an ovule, nor is the 
development of a large amount of sporogenous tissue in the ovule observed 
by Miss Benson in some of the Amentiferae in accordance with our views 
of degeneration. 
1 Avber and Parkin (’07): On the Origin of Angiosperms. Linn. Soc. Journal ot Botany, 
yoI. xxxviii. 
2 Hallier (’08) : fiber Juliania , eine Terebinthaceen-Gattung mit Cupula. Dresden, 1908. 
C C 2 
