20 
R. H. Compton. 
central detached protoxylem is recognised by Miss Thomas, she 
does not subscribe to the view of Sterckx that it is compounded 
partly of a median cotyledonary trace, partly of a root-pole. Her 
view rather approximates to the simple Gerardian conception of 
rotation of half-bundles about a protoxylem pivot, the isolation of 
the latter being regarded as unimportant. The special interest to 
Miss Thomas of the double bundle as an anatomical unit is not so 
much morphological as phylogenetic. She quotes a number of 
cases to demonstrate “ the fairly universal presence of the double 
leaf-trace throughout the Fern-Gymnosperm series, and in a reduced 
form in the Angiosperms, together with its very frequent occurrence 
in modern Ferns.” Its presence in Angiosperm cotyledons is held 
to be a primitive character, illustrative of the principle of onto¬ 
genetic recapitulation, and reminiscent of a fundamental dichotomous 
mode of branching, Such a phylogenetic theory marks a consi¬ 
derable advance on the merely descriptive and analytical views of 
earlier writers on the morphology of the vascular bundles of 
juvenile stages of development. 
The second phylogenetic theory that has emerged from the 
study of spermophytic hypocotyls is due to Chauveaud, to whom 
we owe a series of papers during the last ten years, summarized 
and largely extended in an important work of 325 pages published 
in 1911. 
Chauveaud promulgates a classification of vascular structures 
according to the relative positions of vessels and sieve-tubes. His 
view differs from the stelar theory of Van Teighem and numerous 
English and American authors, and from the theory of the divergeant 
of Bertrand, Cornaille and Chodat, in that both of these theories 
deal with tissues or groups of cells considered as plastic units, 
while Chauveaud treats of the direction of differentiation of the 
characteristic elements of which vascular tissues are composed— 
vessels and sieve-tubes—relatively to one another and to the 
symmetry of the plant-member. His is not a stelar-theory nor a 
bundle-theory, but a vascular-element-theory : it involves a closer 
and more particular—almost histological—study of tissues than do 
previous generalisations on vascular evolution, with the partial 
exception of the above-mentioned scheme of Bonnier. 
Chauveaud’s system is intended to apply to the whole of the 
vascular plants, but for our present purpose only three of his types 
need be considered. These are called respectively the alternate, 
the intermediate and the superposed. In the “ disposition alterne” 
