740 PROFESSOR W. C. WILLIAMSON AND DR. D. H. SCOTT ON THE 
Both cambium and secondary phloém are often preserved in great perfection (see 
photographs 13 and 14, also Witt1amson, Part VII., Plate 6, fig. 31), The phloém 
is essentially similar to that in the stem. The position of the primary phloém groups 
can still be clearly recognized even in advanced roots (see photographs 13 and 14, 
also WILLIAMSON, Part XIII., Plate 23, fig. 21). 
We find no distinct formation of periderm in the older roots, though sometimes a 
few tangential divisions took place in the pericycle (see photograph 14). The 
primary cortex became a good deal compressed by the secondary growth within it, 
but, so far as our specimens show, it was not thrown off. 
Small rootlets, as well as the larger roots, were evidently capable of considerable 
secondary growth. Thus, the most advanced specimen figured (photograph 14) 
can have had only a small tetrarch cylinder to start with, and must in its primary 
condition have been considerably smaller than the other roots figured. 
4. Branching of the Root. 
Specimens showing branches in connection with the principal axis have been 
previously figured.* Additional specimens are represented in our photographs 
10, 12, and 15. The ramification was evidently abundant. Thus, the specimen 
shown in photograph 12 was giving off four branches, almost at the same level, two 
of which are shown in approximately median section. The branches were always of 
considerably smaller size than the main axis. They were evidently rootlets, borne on 
a relatively main root. Two points of interest arise in connection with these 
rootlets :— 
1, Each rootlet is invariably placed exactly opposite one of the protoxylem-groups 
of the main root which bears it (see photograph 12, fig. 20, and the figures above 
cited). This fact is shown with special beauty in the longitudinal section repre- 
sented in photograph 15. Here one of the rootlets is cut exactly in the median 
plane, and its connection with the smallest peripheral xylem-elements of the main 
root is clear. The tracheides of the rootlet bend at a right angle on joining those of 
the main root. The bend was no doubt directed towards the organic base of the 
latter. 
2. The rootlets are evidently of endogenous origin. Although we cannot observe 
their development, this is sufficiently proved by the fact that the rootlet always has a 
cortex of its own, which can be traced through that of the parent root (see photo- 
graphs 12 and 15, also WiLLraMson, Part XIII., Plate 24, fig. 27). 
In every respect, then, the branching of these organs is that characteristic of roots. 
The knowledge which we have now obtained of the roots of Lyginodendron forms a 
striking addition to the characteristics of this extraordinary fossil. We have already 
* Witviamson, “ Organization,” Part VIL, Plate 6, fig. 33; Plate 7, fig. 34; Part XIII., Plate 24, 
fig. 27, 
