306 R. H. Compton. 
(1) both branches may develop more or less equally, forming 
elongated axes which fork again and again: or (2) one of the 
branches may develop fully, while the other, after bearing a very 
few (usually two) pinnae, ends in an aborted bud. This latter is the 
more usual process. 
The dichotomous arrangement is preserved throughout. It is 
very distinct in the forked terminal pinnae ; and it seems probable 
that the appearance of pinnation in the frequent pairing of the 
pinnae at the base of a rachis-segment is a secondary modification. 
This is borne out by a study of the manner in which the vascular 
bundles behave at the fork of the rachis (Fig. 41). The petiolar 
bundle divides in the median plane into two sub-similar portions. 
One of these passes directly into the sympodial continuation of the 
rachis. The other rotates through a right angle, and then gives off 
two successive traces, one of which goes to each of the two pinnae, 
while the central strand disorganises completely in the aborted bud. 
Thus it is clear that the pinnae are branches of the aborting rachis; 
and that they are produced by separate alternating dichotomies 
appears probable. 
Tansley and Lulham (11, p. 478) inclined to the view that in 
M. pectinata the main fork of a frond is a dichotomy in whose angle 
a “ middle lobe ” is developed. This median pinna was compared 
with the bud in the fork of the frond of Gleichenia. They also 
mentioned the alternative view, namely that the “middle lobe” in 
M. pectinata is the inner branch of one of the dichotomies of the 
second order. This latter interpretation appears to fit in with the 
structure of M. sannentosa better than does the former : for in this 
species there is nothing to correspond to the “ middle lobe ” (of the 
first theory), but the branching is a succession of subsequently 
modified dichotomies. The aborted bud, moreover, appears to be 
of different morphological value to that in Gleichenia. 
Thus the comparison with the Dipteris-type appears more apt 
than with the Gleichenia -type : especially as the view which regards 
the latter as a modification of a pinnation appears the most 
natural (see Goebel, 7, p. 319 footnote), and as this view makes its 
connection with a primitive dichotomous frond more remote than is 
contemplated by Diels (6, p. 350) or by Tansley and Lulham 
(11, p.479). 
Root. 
The stele of the root arises alone from the outer cylinder of the 
rhizome, without interrupting its continuity. At the base the 
