the Anatomy of the Cone and Fertile Stem of Eqtcisetum. 453 
That the course of the protoxylem should be irregular is only what we 
should expect to find in a stele undergoing reduction. In so far as the 
elements of the protoxylem are the first tracheides to be differentiated 
in the ontogenetic building up of the stele I think that the protoxylem 
system may well be compared, as it is by Barratt, to a scaffolding. 
She, however, also states that the whole vascular system of the cone is built 
up on this scaffolding, and claims that the protoxylem strands determine the 
main features of the vascular system (Barratt, pp. 222 and 225). If it were 
so the comparison would be an unfortunate one, for a building is constructed 
by means of a scaffolding, but not on a scaffolding, a temporary structure 
that disappears. I do not, of course, mean that the protoxylem is merely a 
temporary scaffolding, for it remains functional though some of its elements 
are destroyed, and undoubtedly plays an important part, since it constitutes 
the first and by 'far the principal source of the vascular supply of the 
sporangiophores. But that the protoxylem strands should alone determine 
the main features of the vascular supply is hardly tenable ; for the develop- 
ment of a parenchymatous mesh above a trace, though occurring later in the 
octogeny, must be associated with the forking or swerving or the dying out 
of the strand of protoxylem, unless the trace is one of those borne on 
a short lateral stalk of protoxylem (see below and p. 454). Both the 
dying out of protoxylem below a mesh and its swerving to one side 
of the latter seem to be derivative conditions, associated with less vigorous 
development of the protoxylem. In the reconstructions of the steles of 
Cone B of E. sylvaticum (Text-fig. 2) and of Cone A of E. debile (Text- 
fig. 4) the course of the protoxylem is indicated by a broken white line. 
Where the growth in thickness has caused the destruction of the protoxylem 
and its replacement by a lacuna the width of the latter is shown by a dotted 
surface. The original elements did not, of course, occupy so wide a space 
as the lacuna. It will be seen that, especially in E. sylvaticum , in which 
the vascular system is better developed, forking of the protoxylem occurs in 
many cases below a mesh. When the protoxylem forks such forking 
obviously occurs before the appearance of a mesh causes the division of the 
metaxylem, and it not infrequently takes place a little below the departure 
of the trace. This is by no means surprising, as when a stele or meristele is 
about to fork the division of its protoxylem often occurs somewhat 
prematurely. When the division of the protoxylem occurs thus early 
the trace often appears as though terminating a short median prolongation 
of the protoxylem between the two forks of the latter. In the case of some 
of the protoxylem strands that pass unforked through the node the trace 
also appears in the reconstructions to be seated on a short stalk, though in 
this case on a lateral one. In such cases the protoxylem of the appendage 
departs from that of the axis a little below the exit of the trace from 
the stele, and, as the sporangiophore to which the trace is destined does not 
