978 Thoday and Ber ridge . — The Anatomy and Morphology of 
use the four-angled ovules of Ephedra altissima to support his view are not 
successful in the light of the anatomical investigations described in the 
current paper ; for so far from it being possible that the outer covering in 
these cases represents only two of the leaves composing the outer covering 
in one of the ovules of the biovulate cones, it is found that it represents the 
fused outer coverings of both ovules, and its four vascular bundles represent 
the six bundles of the two ovules. 
All theories which attempt to explain the outer coverings as composed 
of leaves do not meet the difficulty of the different method of vascular 
supply. Each of the bundles supplying the integument originates as 
the centre one of a group of three bundles formed from one of the bundles 
in the ovular base, the two lateral bundles remaining in the axis (Text- 
figs. II. 11, and III) ; the integumental bundles of Wehvitschia are commonly 
formed in the same manner, but leaf-trace bundles do not arise in this way. 
It appears to us that all attempts to compare the outer covering 
to leaves requires some distortion of the evidence, and that it is more correct 
to regard it as an integument in the ordinary sense of the word — a covering 
of the ovule of problematical origin. This is still more true of the inner 
covering, which with its micropylar tube is obviously comparable with the 
integuments of other Gymnospermous ovules. The complex structure of 
the outer covering is also unlike that of the ordinary leaves and is much more 
comparable, not only with that of the outer covering in W elwitschia and the 
middle one of Gnetum } but with that of the outer covering of other ovules, 
more especially those of the Bennettitales. The resemblance 2 between the 
outer covering of Gnetnm and the integument of Bennettites to which atten- 
tion has already been drawn is still further emphasized in Ephedra ; for here 
the three or four fibrous strands representing the star-like rays of the fibrous 
layer in Gnetnm and Bennettites each .correspond to an external angling 
of the seed, such as is present in Bennettites . 3 The remarkably detailed 
resemblance in the structure of their outer integuments between the three 
members of the Gnetalean alliance, their leaves being so different, is alone 
a striking fact, and suggests that the coverings of the ovule are independent 
structures, less plastic than the leaves. 
The ovule with its two integuments thus remains an isolated structure 
in the axil of the bract. There is no evidence to suggest that the ovule was 
ever borne on the subtending bract, and there is no direct evidence, except 
the analogy between its vascular supply and that of a vegetative bud, that 
1 It is clear from the structure that it is the middle covering of Gnetnm which is homologous 
with the outer covering of the other two forms ; Thoday (Sykes), 1911. 
2 Berridge, 1911 ; Thoday (Sykes), 1911. 
3 Wieland’s recent figures of the ovule in Cycadeoidea turrita (1911, p. 458, Fig. 15 c, and 1912, 
p. 90) still further strengthen the comparisons which have been made between the ovules of Gnetnm 
and Cycadeoidea, though his comparison with Gnetnm, 1911, p. 458, Fig. 15, a and b, is evidently 
based on some misapprehension. 
