l 55 
Lycopodiales. 
largest definitely Lepidodendroid cone with the structure preserved 
known to him is Lepidostrobus Brownii ; this cone is about one 
quarter of the size of the largest Ulodendroid scars (25). Graf 
Solms-Laubach, however, mentions various colossal cones (21); the 
details of their structure are unfortunately not known and they may 
not have belonged to the Lepidodendraceae. In any case, as they 
appear to have possessed an axis of exceptional size, they could not 
very well have been attached to any Ulodendrou- scar known to us, 
for the umbilicus or point of insertion of the supposed axis of the 
cone is invariably small. Mr. Watson points out that the shape of 
the Ulodendroid scars is inconsistent with the view that they 
increased secondarily in size after the cones had been detached. 
He regards the scars as formed by the attachment of branches and 
considers the umbilicus as the point of the stele of the branch ; he 
asserts that this theory is supported by the arrangement of the 
markings on the scar itself, or at least on its lower part; these 
resemble the leaf-traces of the order. The unequal size of these 
dots, which it would be difficult to account for if they are regarded, 
as is usually done, as leaf bases, is explained by the fact that some 
of these departing leaf-traces would naturally be cut through 
tangentially. The similarity of these markings to Lepidodendraceous 
leaf-traces had already been commented on by Graf Solms-Laubach 
(21). Mr. Watson contends that the ill-defined structures, rarely 
found attached to the Ulodendroid scars, are not, as had been 
been supposed, cones. It seems very likely that his theory, if not 
completely satisfactory, yet comes nearer to the truth than the 
commonly accepted interpretation of the facts. It has the advan 
tage of not supposing that such an ancient order as the Lepido- 
dendraceae had acquired a character, namely the production of 
cones on old stems, which on comparative grounds, and on account 
of its uniqueness, was probably not primitive. On the other hand 
Dr. Scott, in the final chapter of the first edition of his “Studies,” 
remarks that the Palaeozoic Lycopods were more specialized than 
their recent allies in the sharper definition of their strobili ( Pina - 
kodendron being then unknown) and he adds: “The frequent 
differentiation of the shoots on which the cones were borne ( e.g ., 
the Halonial branches of Lepidophloios) is another indication of the 
same rule” (13). These Halonial branches seem to have been the 
ultimate ramifications of Lepidodendroid stems, and the scars 
which they bear were thought to be of the same nature as those of 
Ulodendrou, though they were much smaller and usually spirally 
