239 
for Chelone-and Digitalis both have the same structure, viz: only 
tracheze and libriform, but in these cases the stems examined 
were of only one year’s growth and the woody portion not well 
developed ; the specimen of Veronica Andersoni on the contrary, 
was a fair example of woody development. In Aalleria, 
Celsia, Scrophularia and Scoparia, in addition to these elements, 
medullary-ray cells occur, but no wood parenchyma. In Paulow- 
uta and Diplacus the wood parenchyma is added; in Paulownia 
this surrounds the large trachea, nosmall ones occur. In Dép/a- 
cus the larger trachez lie in radial rows along the medullary 
rays, thus in both instances the tracheze are in connection with 
living, acting cells. In Freylinta the libriform contains both 
bordered and simple pores, the former predominating ; all the 
other elements belonging to the most complex type are also found 
here, unless it may be the thin-walled tracheids, as it is not pos- 
sible always to decide between these and the small trachea. So 
here we have a most marked contrast to the simple structure of 
the woody tissue of Veronica. 
This variation of structure in the genera of a few families pre- 
sents a striking contrast to the uniformity prevailing in the great 
majority. While it suggests at once the idea that the degree of 
relationship between the members of those few families may be 
very remote, still in those instances where this idea is not sub- 
stantiated by corresponding morphological differences, as in case 
of the last family described, the possibility of transition stages in 
the development of the border naturally suggests itself. 
Assuming only what is self-evident in reference to the func- 
tion of the pores, that is, that they are designed to allow a freer 
communication between the cells than the walls otherwise fur- 
nish, we may suppose that in case of increased demand in this re- 
spect there might gradually occur a corresponding change in the 
form, size, or frequency of the pores to meet this demand. 
First, as regards the form, it is possible that in this way the 
simple pore with its funnel-shaped enlargement at the base may, 
by a succession of changes, have developed into the bordered 
pore, the curve of the enlargement having grown more and | 
more defined until at last it has changed into a decided angle. 
The attempt to verify this supposition by observation is, of - 
