Vascular Anatomy of Tubers of Nephrolepis. 79 
also the strands diverge widely from the base and converge again 
towards the apex ( loc. cit. p. 41, and Plate I, fig. 3). Tubers similar 
to those of Nephrolepis have also been recorded by Ule 2 in a species 
of Hymenophyllum, hut unfortunately nothing more is known about 
these. 
Since the behaviour of the vascular strands in the tubers of 
Nephrolepis cannot he explained as being due to any influence from 
leaf-traces, we must look for an answer to the problem elsewhere. 
For a plausible suggestion, which would apply equally well to the 
case of Equisetum, I am indebted to Mr. R. H. Compton who 
believes that an increase in the diameter of an organ tends to cause 
a corresponding dilatation of its vascular system. As he has pointed 
out in another connexion, 3 “ the impression given.is that the 
stele takes the opportunity afforded by the increased diameter to 
acquire a pith and expand.” While dealing with the factors in the 
evolution of solenostely Mr. Tansley 4 has remarked that an increase 
in the diameter of the stele would be necessitated by any broaden¬ 
ing in the span of the C-shaped leaf-trace, to the attachment of 
which on an originally solid stele he has ingeniously traced the root- 
cause of the departure towards solenostely. In the case of 
Nephrolepis and Equisetum the great increase in the diameter of the 
axis may exert on the stele an influence similar in effect to that of 
the enlarging leaf-trace on the cauline stele. For the proper supply 
of food and water to a storage organ of large diameter, it would be 
necessary that the vascular system should not have the form of a 
solid axial cylinder, but of an expanded hollow stele. 
The further change, from the tubular to the net-like condition 
may be taken to be a natural result of the enormous dilatation of the 
stele, and the concomitant thinning of the tube in places to produce 
the “ perforations.” Contrary to what is the case in the normal 
fern rhizome, where the dilatation of the stele is, comparatively 
speaking, slight, we may confidently believe that the mere thinning 
of the tubular stele would suffice to produce gaps in it. In fact, 
a suggestion of such a mode of origin for these perforations is visible 
in their irregular shape, in the frequent occurrence, at their edges, 
of thin blindly-ending strands, and in the varying width of the 
1 Duval-Jouve, “ Histoire naturelle des Equisetum de France,” 1864, p. 6 
ff. ; Luerssen in Rabenhorst’s Kryptogamen-Flora, iii., 1889, figs. 203, 206, 
209. 
* Ule, E., Berichte d. d. bot. Ges., XV., 1897, p. (85). 
3 Compton, R. H., “ An Investigation of Seedling Structure in the 
Leguminosae,” Journ. Linn. Soc. 1912, p. 97. 
4 Tansley, “ Lectures on the Evolution of the Filicinean Vascular System,” 
New Phytologist, 1907, p. 150. 
