No. 464] STUDY_OF THE SALICACEE. 533 
a dwarfed or even prostrate habit of growth no longer imposes 
those demands for rigid support which are from the first a prime 
necessity of the strictly arborescent forms. Precisely such con- 
ditions and precisely such results are to be met with in Salzr 
cutleri Tuckerm., of which a more detailed account will appear 
on a subsequent page. 
In the development of the secondary xylem of angiosperms, 
two distinct lines of evolution are to be observed —the one lead- 
ing to the exaggerated development of vessels, the other leading 
to the exaggerated development of mechanical tissue, the two 
being balanced with respect to the relative excess of require- 
ments on the one side or the other. Confining our attention in 
the first instance to the vessels, it is found that in their evolu- 
tion from the elements of the protoxylem, these latter deviate 
more widely from a fibrous form so that their products become 
broader and either actually or relatively shorter, while the termi- 
nal walls often disappear altogether, thus giving rise to tubular 
structures of indeterminate length. The walls of such vessels 
are somewhat modified by secondary thickening, though relatively 
to their usually great breadth this is not more than is demanded 
by their own need for support, and it cannot be regarded as an 
essential factor in the strength of the organism asa whole. Like 
the tracheids of the Coniferae, the walls are characterized by the 
presence of bordered pits which differ in two very important 
respects from the similar markings which generally characterize 
the higher gymnosperms. They are always multiseriate and 
chiefly hexagonal, though as shown in the Salicaceze, they may 
become more distant and definitely rounded ; they also occur on 
both the radial and tangential walls. In both of these features 
the vessels show the survival of primitive characteristics which 
are only partially represented among the Cordaitales, and which 
indicate that the ancestral forms must be looked for among the 
Cycadofilices where they are of well known occurrence, or else 
among some other group of similar structural characteristics, 
but of which we have no present knowledge. Now the angio- 
sperms are essentially all broad-leaved plants adapted to a dis- 
tinctly mesophytic habit, and it is therefore to be concluded that 
the transpiration current not only moves with greater rapidity 
