700 
Thomas. — Seedling Anatomy of 
structure is by no means the same as that of the root. When the hypo- 
cotyledonary region is long enough for its structure to appear uninfluenced 
by the changes associated with the cotyledonary node and collet respec- 
tively, it is seen to consist essentially of a more or less pithed xylem mass 
with a varying number of practically exarch protoxylem groups, the whole 
sometimes almost completely surrounded by phloem, but more often with 
the phloem broken into groups alternating with the protoxylem groups. 
All the conducting tissue present at a very early age is directly con- 
tinuous with that in the cotyledons in an upward direction, and with that in 
the primary root in a downward direction. 
In Text-fig. 42 (p. 730) the phloem breaks away along the line indicated, 
and a xylem group, together with the two phloem groups thus produced, 
passes into the base of one cotyledon ( a , b ), constituting the curious 4 double 
bundle ’ described above, the other cotyledon being supplied in a similar 
manner. At the other end of the hypocotyl the extended phloem groups 
of the latter are continued by the much smaller and more compact ones of 
the root (e), while the rather scattered xylem is continuous with the definite 
diarch plate of the same, the pith having disappeared. The ill-defined 
endodermis of the hypocotyl passes into the well-defined endodermis of 
the root. 
In Text-fig. 43 (p. 731) a similar process takes place, but in addition to 
the central protoxylem of the 4 double bundle ’, groups of protoxylem on the 
flanks continue the intercotyledonary groups of the hypocotyl ( a , b). (See 
also PI. LI, Figs. 24, 25 of Althaea rosea) In other forms these gfoups are 
separated from the double strand as distinct laterals, e. g. Magnolia tripetala , 
and the double bundle itself may be represented by what appears at first 
sight, particularly at some ages and at some levels, to be two perfectly 
distinct strands. (See Text-fig. 38, p. 722, of Prunus persica.) Following 
the strands downwards through the hypocotyl, the intercotyledonary proto- 
xylem groups are seen to pass into the intercotyledonary poles of the 
tetrarch root {c). 
The most marked 4 transition ’ takes place somewhere in the cotyledon, 
where the double bundle becomes continued upwards as a single collateral 
strand. Most authors speak of this as the delayed fusion or precocious 
splitting of phloem groups, according to direction of description, but none, 
so far as I am aware, have suggested any reason for the delay, or, in other 
words, for preparation for root formation in the cotyledons. The double 
bundle, moreover, is met with in its most extreme form in those species 
in which root structure is furthest from the cotyledonary node (Euphorbia- 
ceae, Calycanthaceae). The homology with the double bundle of the 
widely separated central strands of these forms is proved by their union in 
the upper part of the cotyledon, and by the identification of isolated central 
protoxylem elements in the lower part. 
