Stigmarian Rootlets. 565 
central cylinder as it does in Fig. 5 of the present communica- 
tion, and not at right angles as was the case with the rootlet 
figured by Renault and that figured previously by me (’ 02 ). 
The correctness of Dr. Scott’s view is further borne out by 
several rootlets in the Manchester Museum, one of which is 
represented in Figs. 3 and 4 1 . This rootlet is one of a type 
not uncommon among Stigmarian rootlets, but differing both 
from the one first described by me with a vascular branch 
(’ 02 ), and also from the three rootlets described above. It 
will be seen that ‘in this third type of rootlet the stele does 
not lie freely in the space left between the inner and outer 
cortex, but is definitely connected with the latter by a strand 
of parenchymatous cells, reminding one of the similar con- 
necting strand in the roots of Isoetes. Such rootlets have been 
figured before, but no special attention has been drawn to 
this feature, which might from Williamson’s Fig. 16 2 be 
thought to be an attribute of very young roots only. A 
careful examination of a large number of rootlets of this type, 
many of them evidently fully developed, convinces me that 
it is a somewhat different type of rootlet. In the one repre- 
sented in Fig. 3 it will be seen that this parenchymatous con- 
nexion between the stele and the outer cortex is traversed by 
three strands of tracheids (tr\ tr\ tr'") similar to those figured 
by Williamson, and rightly interpreted by Dr. Scott as vas- 
cular branches similar to those described by Renault. From 
the number of these isolated groups of tracheids it would 
appear that such branches were given off at short intervals 
in the rootlets of his type, and that they ran very obliquely 
towards the cortex, so that two or three strands are cut across 
in transverse section. Their oblique path is evident from the 
enlarged drawing of a portion of this rootlet (Fig. 4), in which 
the lateral markings of the tracheids are visible in a trans- 
verse section. It is, of course, possible that these strands were 
not all of independent origin, but that the appearance may be 
due to some branching of the strand similar to that of Fig. 5 
1 Hick Collection, No. 75. 
2 Williamson, W. C. (’ 81 ), PI. LIII, Fig. 16. 
Q q % 
