216 Jeffrey.— On the Structure of the Leaf in Cretaceous Pines. 
Island differed from modern Pines, (i) in the better development of the 
transfusion elements around the bundle ; (2) in the differentiation of the 
transfusion elements into an inner sheath composed of elongated tracheidal 
elements and an outer, much broader zone made up of more nearly isodia- 
metric elements with thinner walls ; (3) in the probable absence of an 
endodermis ; and (4) in the absence of infolding of the walls of the mesophyll. 
They resembled their modern descendants in possessing a fixed small 
number of verticillate fascicular leaves for the brachyblasts of each species. 
The growing-point of the brachyblast in the true Pines of the Cretaceous 
often persisted, although the author in no case has found it to be as promi- 
nent as in Preplans , The wood of the leaf-bundles was entirely endarch. 
Prepvius statenensis differed from other Cretaceous Pine-like Conifers in the 
possession of true centripetal wood, such as has been described in the leaves 
of no other Conifer living or extinct. The fascicular leaves were attached 
in an indefinite large number spirally to the brachyblast, which was deciduous 
as in the true Pines, but differed from them in having a prominent persistent 
growing-point covered with protective scales. The transfusion tissue of 
Prepmus has the same general organization as in Cretaceous species of 
Pinus , but it was much more abundant, and the outer zone had no admixture 
of parenchymatous elements. 
Conclusions. 
The presence of centripetal wood in Prepinus in the present state 
of our knowledge can scarcely be interpreted in any other way than as the 
persistence of an ancestral character possessed by the parent stock of the 
Abietineae. If it be conceded that the centripetal wood which is found 
in the leaves of Prepinus is a vestigial feature, it cannot well be denied that 
the double transfusionary sheath accompanying it in that genus and still 
retained in the Middle Cretaceous representatives of the true Pines is sus- 
ceptible of a similar interpretation. This view of the matter is rendered 
still more probable because of the close agreement of the presumably 
primitive Abietineous leaf-bundle with that found in certain Cordaites 
which are regarded, by those whose studies fit them to judge, as the 
ancestral stock from which the Coniferales have been derived. If we may 
venture to depict the phylogenetic development of the Coniferous foliar 
bundle in the light of the new facts supplied by the study of their Mesozoic 
representatives, it would appear to be somewhat as follows. The ancestral 
Conifers, closely allied to the Cordaitales, possessed foliar bundles charac- 
terized by the presence of centripetal xylem and transfusion tissue of 
a complex type, consisting of a double cordon, an inner sheath composed 
of elongated tracheidal elements, and an outer jacket of ordinary transfusion 
cells without any admixture of parenchyma. In the course of subsequent 
