A Theory of the Double Leaf-Trace. 9 ! 
soon the dichotomy becomes unequal, so that a more or less 
completely pinnate form of branching takes its place. At the same 
time a definite petiole develops, so that the region of precocious 
dichotomy of the vascular system is much drawn out. This results 
in the presence of two strands throughout the petiole. In some 
cases a single strand is reformed, by the fusion in the petiole of the 
originally separate bundles, which open out again when the lamina 
is reached. 
Potonie 1 has drawn attention to the repeatedly dichotomous 
mode of branching of the fronds in Palaeozoic impressions, and 
Oliver and Arber 2 agree that this is probably indicative of the foliage 
of the Cycado-Filices rather than of true Ferns. If this be so, it 
affords a ready explanation of the more obvious presence of the 
double trace in the Cycado-Filices than in the Ferns. The common 
occurrence of fusion in the petiole of Lyginodendron, of the 
originally separate strands of the leaf-trace, is very plausibly 
explained by supposing that the primitive early dichotomy of the 
leaf is delayed by the differentiation of a petiole. The similar fusion 
found in the early leaves of some Ferns has been referred to. 
Dichotomy may then perhaps be considered the mode of foliar 
branching which probably obtained in the early fern-like groups from 
which the Cycado-Filices and the Ferns arose. Its influence 
is seen in the irregular dichotomy of the fronds of the Cycado- 
Filices, and in the dichotomy of the early leaves of modern 
Ferns. It may perhaps be connected with the splitting of the 
cotyledons, which is one of the causes of the polycotyledony of 
the Conifers, and which is also occasionally found in other groups 
It probably also directly or indirectly accounts for the double leaf- 
trace of the Fern and Fern-Gymnosperm groups, and it finds its 
latest expression in the “ double bundle ” of the cotyledons of 
Angiosperms. 
In conclusion I wish to express my great indebtedness to Mr. 
Tansley for his generosity in suggesting that I should publish these 
results from our joint work. It is impossible to estimate the 
influence which the benefit of repeated discussions with him has 
had on the theoretical views here brought forward. My thanks are 
also due to him for much helpful criticism during the preparation of 
this paper. 
It is hoped that a much more exhaustive and comprehensive 
account of seedling anatomy will be possible before long. 
1 H. Potonid. Ber. deutsch. bot. Gez.XIII., 1895, etc. 
2 E. A. N. Arber. “On the Past History of the Ferns,” Annals 
of Botany, 1906. 
