814 
THE AMERICAN NATURALIST. 
[Vor. XXXIX. 
referred to as (1) and (2), and their more detailed character- 
istics will be discussed in connection with the 
radial section. In the genus Populus, as clearly 
shown in P. tremuloides (Figs. 12, 13), there is 
no clearly defined distinction between the two 
kinds of cells by means of which each may be 
recognized beyond question. In Fig. 13, the 
distinction is clear by reason of the distribution 
of the resin and the special location of each kind 
of cell. Thus the more resinous cells, 2, occu- 
pying the central region, are of the first order 
(1) while the less resinous cells, 4, occupying 
the extremities of the ray, are of the second 
order (2). This differentiation serves to illus- 
trate the general law that the cells of the first 
order are usually the more strongly resinous, 
and that the less resinous cells 
of the second order are chiefly 
Fic. 12. — Populus . . 
tremuloides. Tan- terminal unless also interspersed. 
gential section of But such distinctions are not al- 
the narrow medul- ü F 
ways valid, even in the same spe- 
‘œ Cies, since in Fig. 12, taken from 
non-resin- 1 
ous cells ofthe sec- the same section, the non-res- 
ond order. X 350, . ^ š 
3 7^  jnous terminal and interspersed 
cells, 2, are probably all of the second order, 
while the more resinous cells, a, are chiefly of 
the first order. But the three resinous and ter- 
minal cells of this figure involve considerable 
doubt as to their precise character, inasmuch 
as their positions would lead us to assume that 
they must be of the second order in spite of 
their resinous contents, à view which gains force 
from what may sometimes be observed in radial 
section. 

In the genus Salix (Figs. 14 and 15), where 
the structure is of a more 
Character, the cells of the 
usually all resinous, 
decidedly resinous 
medullary rays are 
and this feature serves no 
useful purpose in differentiations. We do find, 

the second order. 
X 350. 
nevertheless, 
