414 MISS T. L. PRANKERD ON THE STRUCTURE OF 
bundles in Physostoma, хо that the true dimensions of the latter may be 
somewhat greater ; and, further, that the limits of fossil vascular strands are 
not always sharply defined, and might be interpreted differently by different 
observers. 
Histologieally the bundles are composed of six or eight tracheids with a 
few secretory sacs and. thin-walled elements, probably parenchyma, of much 
smaller dimensions than the cells of the soft tissue around (РІ. 22. fig. 2 
and Р]. 28. fig. 20). It is only occasionally, however, that even this much 
detail is evident, for the bundles are often dragged from their normal 
position, and sometimes more or less obliterated, or absent altogether. The 
tracheids were scalariform and about 15 ш in diameter; it seems very doubtful 
if there was any definite protoxylem. So far аз can be made out, slightly 
smaller elements occur in any position, but, in view of the fact that smaller 
elements are not necessarily protoxylem, and the impossibility of determining 
the real sequenee of formation, we are of necessity left very much in the 
dark as to the nature of the bundle in this respect. In lt. 48 (1) a bundle 
in longitudinal section seems to show a smaller internal element (РІ. 22. 
fig. 4), but it is difficult to say that this is not merely а tracheid of equal 
size cut tangentially. One of the bundles in R. 41 (1) shows on its inner 
side what may be interpreted as a loose spiral, but is, by itself, scarcely 
convincing. 
So far as any reliance is placed оп the above, it will of course point to 
the endarch nature of the bundle. Towards the apex of the seed the bundles 
thin out to two or three elements, and apparently die out at about the level 
where the ribs end and the ridges begin. 
А word more may he added with regard to the apex, before passing to ihe 
detailed description of the body of the seed. As the apex is approached, the 
bundles and fibres diminish, the former appearing to die out before the latter, 
though it is impossible to give ап exact deseription of the point without the 
appropriate section. Bundles apparently die out in. all sections, but this is 
of course due to obliquity—the plane of the section and that of. the bundle 
never exactly coinciding. Owing to this, and also to the frequent displacement 
of the bundles, опе could not easily be certain that the ending of the 
bundle was present, even if that were actually the ease. 
Near the apex, the tips of the loeuli, bounded by the inner and outer 
epidermis, seem to contain soft tissue, though apparently thieker-walled 
than the corresponding tissue in other parts of the seed (РІ. 83. fig. 17). 
Secretory sacs are present, but in no section passing through this region does 
there seem to be any indication of hydathodes, nor is there any differentiation 
of the apical epidermal cells, such as a thinning of the cuticle, which would 
lend any support to ihe idea that the apices of the canopy chambers were 
porous, as was once suggested (13, p. 462). 
