818 THE AMERICAN NATURALIST. (VoL. XXXIX. 
Together with the cells of the first order, cells of a somewhat 
and often wholly different characterappear. In the genus Popu- 
lus, this differentiation is expressed with the least force, and it 
is difficult or even impossible to distinguish between the two 
except through certain special structural features of the radial 
walls. Such cells are, generally speak- 
ing the least resinous. To all such I 
have applied the designation “cells of 
the second order" or (2). An inspec- 
tion of Fig. 16 taken from Populus 
tremuloides, will show that there is es- 
en — Populus tremuloides. sentially no distinction as to form and 
section of cells of the sec- 
ond order (2) showing oval pits on Structure, between cells of the first 
bes walls opposite vessels. order g, and those of the second order, 
6, and it is only when we study such 
cells opposite vessels that the distinction becomes obvious. In 
such regions cells (2) invariably show either oval (Fig. 18) or 
quadrangular pits (Fig. 19) disposed in somewhat definite radial 
series. Such pits and of the forms described, are characteristic 
of the genus Populus, and from this there is no essential devia- 
tion so far as our investigations have pro- 
ceeded. These features make it clear 


cdm Salix the case is very different, Ftc- 19.— Populus tremuloides. 
ere we alw i Radial section of cells of the 
teri ays find the Tay charac- . second order, (s) ehowiig quede 
erized by the presence of cells which rangular pits on the lateral walls 
opposite vessels. X 350- 
ay that the cells become 
and therefore high and narrow. 
alix, and it has not been observed 
It is this very strong difference in 
differentiati 
lon between the two kinds of cells as exposed in tan- 

