32 Mitchell.—Contributions towards a Knoiv ledge of the 
Round the outer edges of the phloem a few protophloem elements may 
still persist. There are one to two layers of phloem parenchyma. 
e. S. Martensii. 
The pericycle is one layer deep. Protophloem elements are rare. 
Generally one layer of sieve-tubes occurs (sometimes there are two), and 
these may be entirely absent opposite the protoxylems. The phloem 
parenchyma is somewhat elongated. 1 
The Ligide . 
In a previous paper Professor Harvey-Gibson ( 14 ) has given a detailed 
account of the structure and development of the ligule in Selaginella , 
so that to describe in detail the ligule of the cone would be merely 
to repeat the facts recorded by him. In brief, it may be said that the 
ligule of the cone, down to the minutest details of glossopodium, glosso- 
podial sheath, &c., has essentially the same structure as that in the 
vegetative axes ; where a tracheal cup is present in the latter, as in 
.S', helvetica , 5 . inaequalifolia (PI. Ill, Fig. 5), &c. f its counterpart is seen 
in the cone. As a general rule, the ligule of the cone is larger and more 
expanded than that in the stem. It always persists in a fresh condition 
quite down to the base of the strobilus, even when the ligules just below the 
cone are shrivelled and dead. (These facts are well exemplified in 
A. Kraussiana , 6'. inaequalifolia , &c.) 
The cuticularization of the cells radiating out from the glossopodium 
to the vascular system does not take place in the cone until the spores have 
been shed (cf. 14 ). Fig. 13,‘PI. IV, is a longitudinal section through the 
ligule of 5 . Wildenowii —in this case all the sporangia were empty. 
The ligule commences its development early, and matures much more 
rapidly than the sporangium which it subtends. This, taken in conjunction 
with the above facts, suggests that the ligule is closely connected in function 
with the sporangium ; the two commence their development together, and 
the ligule becomes functionless as soon as the sporangium has shed its 
spores. The proximity of the ligule to the vascular supply, and the 
mucilaginous nature of the cells of its lamina, which are, therefore, specially 
adapted for storing water, suggest that its function is to prevent the drying 
of the sporangium during the development of the spores ( 14 ). 
The examination of the above species serves to emphasize the inter¬ 
mediate position in which Selaginella stands. On the one hand, it may 
revert to conditions characteristic of the distinctly primitive Lycopodia, 
1 It seems worth while drawing attention to the simplification of the vascular complexities of 
the stem in such types as S. Lyallii and S. i7iaeqiialifolia as the cone is approached, which suggests 
parallel arrangements seen in such forms as the Cycadaceae. The cone would appear to be more 
conservative of the primitive vascular state, and the simplification of anatomical structure might be 
regarded as evidence of the independence of evolution of the two regions, each on its own account 
when once fairly segregated. I am indebted to Professor Bower for this suggestion. 
