35 6 
Notes . 
v 
ON THE RELATION BETWEEN TRIGONOCARPUS AND GINKGO.— 
Miss Affourtit and Miss La Riviere, in their paper entitled £ On the Ribbing of 
Seeds of Ginkgo’ (Ann. Bot., October, 1915), writing of the comparison of Ginkgo 
with Trig onocar pus , stale (p. 594) that ‘ Since in Ginkgo , however, no valves occur — 
the stony coat lacking fissures at the plane of the ribs — and as vascular bundles are 
absent from the sarcotesta, those seeds cannot, as it seems to us, be compared with 
the seeds here described \ With regard to the first objection, it has been pointed out 
by the writer (Ann. Bot., January, 1914) that the Trigonocarpales show almost every 
transition with regard to the occurrence of commissural ribs. In Polylophospermum 
both the major and secondary ribs were commissured. In the genus Trigonocarpus 
itself the fissured character had entirely disappeared from the secondary ribs and was 
not uniformly exhibited by the major ones. Moreover, in the closely related Stephano- 
spermum both ribs and commissures have entirely disappeared. The absence of 
fissures in the major ribs of Ginkgo is therefore merely a further stage in the 
evolutionary tendency exhibited by the genus Trigonocarpus , and it is significant 
that, as pointed out by Carothers (Bot. Gaz., 1907, p. 126), the integument of Ginkgo 
readily splits in the plane of the ribs. 
The absence of sarcotestal bundles in Ginkgo can no more be taken as 
precluding affinity between the two groups than the presence of vascular strands 
in the integuments of some angiospermous ovules invalidates their comparison with 
ovules in which an integumental vascular system is lacking. 
In view of the absence of sarcotestal bundles, the non-development of tertiary 
ribs calls for no explanation. 
It is probable that the two vascular bundles (three in three-angled seeds) of 
the Ginkgo ovule correspond to the nucellar supply of Trigonocarpus. 
This is indicated by the facts that they pass up close to the plane of fusion 
between nucellus and integument, and that, though serving as the vascular supply for 
both structures, the bundles end at the level at which the nucellus becomes free. 
That the number and position of the strands should correspond to that of 
the ribs of the integument is not surprising, seeing that the angling of the nucellus 
shows a like correspondence as to number and position. Moreover, in Trigonocarpus 
shorensis it was found that the number of bundles in the nucellar system was a 
multiple of three, corresponding with the trimerous character of the integument. 
In the taxonomically more important features of general organization the ovules 
of Ginkgoales, Cycadales, and Trigonocarpales exhibit a uniformity of construction 
difficult to explain except on the basis of affinity. The morphological and anatomical 
characters of these groups, whilst emphasizing the closer relation between the 
Trigonocarps and Cy cads, lend further support to the hypothesis of the affinity of all 
three. On such a view, the large proportion of Ginkgo ovules with three ribs recorded 
by Affourtit and La Riviere has an added significance. 
E. J. SALISBURY. 
East London College, 
January , 1916. 
