Vascular Anatomy of Tubers of Nephrolepis. 77 
by many fern-rhizomes during their individual development from the 
protostelic to the dictyostelic condition. Fig. 3, A, for example, is 
strongly suggestive of the “ Lindsaya-type ” of stele, while Fig. 2, 
A is exactly the appearance presented by a transverse section 
through the internode of a typical solenostelic fern. Neglecting for 
a moment the mode of origin of the gaps, the appearance of a 
transverse section through the middle of the tuber does not 
materially differ from that of a dictyostelic rhizome. 
Roots arise promiscuously on the tuber, hut sometimes they 
are almost confined to the narrow basal end and to the apical 
mamelon. As their strands are sometimes attached to the net-like 
stele at the base of a gap, sections through those regions (Fig. 2, B, 
C) give the impression that the gap is “ caused ” by the exit of the 
root-traces. An examination of Fig. 1, A will, however, show that 
the point of origin of the root-strands does not follow any rule. 
It seems probable that the relative duration of the different 
stages in the transition depends largely upon the shape of the tuber. 
If the increase of diameter from the stolon to the tuber is gradual 
(as shown in Fig. 1, A) the expansion of the vascular system is 
correspondingly slow, and the intervening tubular stage, with 
internal phloem, pericycle and endodermis, is of appreciable length. 
If, on the contrary, the tuber is sharply marked off from the stolon 
the different stages in the transition succeed each other rapidly and 
may even overlap, so that the xylem-ring may become disintegrated 
before it has acquired an internal pericycle, endodermis and ground- 
tissue (Fig. 3, C). In the light of this suggestion some of the 
individual variations met with in the mode of transition may 
perhaps be satisfactorily explained. This view appears to find 
support in the fact that in the pear-shaped tubers examined the 
transition was more rapid at the apex than at the base (that is, the 
narrower end). 
Professor Heinricher ( loc . cit., p. 71) has found that the shape 
of the tuber seems to be characteristic for each species and suggests 
that it may be a character of taxonomic value. Thus in N. hirsutula 
Presl apud Raciborski the tubers were pyriform, while in specimens 
of N. tuberosn (Bory, Willd.) Presl=N. cordifolia (L.) Presl they 
were found to be ellipsoid. In view of the confused condition of 
the taxonomy of the genus such a distinction would be welcome, 
but it is difficult to say what importance can be assigned to this 
character, for variations in the shapes of tubers from the same plant 
are not rare. In a few tubers from a specimen of the last-named 
species a complete transition from the pyriform to the ellipsoid 
