1886. | BOTANICAL GAZETTE. 259 
ducts, or central. The term “ hypoderma” still has its use, how- 
ever, as it contains another group of cells which we have called 
“thin-walled cells,” to distinguish them from the strengthening 
cells, This thin-walled layer frequently occurs between the epi- 
dermis and the cortical strengthening cells, and its presence or 
absence is a character of considerable importance. Seven of our 
species have an hypoderma composed only of these thin-walled 
cells, with no cortical strengthening cells; eight or nine species 
have no thin-walled layer between the epidermis and the cortical 
strengthening cells; while the remainder have the thin-walled 
layer between the epidermis and strengthening cells. By “ thin- 
walled cells” it must be understood that we are speaking com- 
paratively, as they are by no means thin-walled in fact, but ¢on- 
trasted with the epidermal and strengthening cells are decidedly 
so. Rarely is there any difficulty in distinguishing this layer, but 
occasionally, as in P. monticola, the thin-walled layer shades into 
the strengthening cells. 
Il. The mesophyll region. This is chiefly compused of large 
chlorophyll-bearing parenchyma cells, with very characteristic 
infoldings, which are of no diagnostic value. In this region, 
however, are found the resin ducts, and their position and size 
furnish very important characters They are found in three 
positions, viz.: peripheral, when they lie next to the cortical 
region; parenchymatous, when completely surrounded by the 
mesophyll ; internal, when next to the bundle-sheath. The terms 
“ external,” “ medial,” and “internal” would better express their 
that in two species resin ducts have been occasionally discovered 
in the fibro-vascular region, viz.: P. sylvestris,’ to which we a 
P. serotina. Dr. Engelmann considers the positions of these ducts 
in the mesophyll region as the most useful diagnostic character 
obtained from the leaf-structure. However, even this character 
can not be relied upon exclusively, as variations from the normal 
changing the normal position, but in the development of accessory 
ducts in some other position, or in the change in position of a 
single one of the normal ducts. To Dr. Engelmann’s list of these 
variations we have added ten or twelve species, showing that such 
variation is not unlikely in the whole genus. All the resin-ducts 
2Arthur, Barnes and Coulter, Hand-book of plant dissection, p. 167, 
