328 BOTANICAL GAZETTE [NovEMnER 
section. The difference then between the inner and outer face 
of a ring of growth lies in the more armed shape of the inner — 
cells, and the less armed shape of the outer, and the optical 
characters of the more-armed cells in the tissue are rendered | 
peculiar by their tendency to stand in radial rows for some dit 
tance at the same level. In general the outer rings show longer 
rays and a generally coarser structure than do the inner, while 
the innermost ring of all around the pith shows the rays bet 
sparingly, and in cross section is almost everywhere strongly 
sclerenchymatous in aspect. In the outer zones I have seen 2 
many as nineteen cells in one of the rays, showing that this 
number of armed cells stood at exactly the same level ina radial 
plane and were all cut directly through the middle when the 
section was taken. A comparison of the cross and long sectioms 
as presented in the plate should make this matter clear. None 
_of the zones of growth, except the one immediately around * ; 
pith, are free over any considerable area from the rays resulting 
in the manner described. Lying outside of the outermost - 
will be found the immediate products of the superficial cambium, 
not yet modified into the characteristic armed-cells, but present: 
ing a more isodiametrical character. The actual cambial pete 
seems to lie immediately under the epidermis. Th 
cortical region and extending down to the region of armed § 
renchymatous cells chloroplastids are abundant, and the ep 
mal cells themselves, with the layers immedia 
dense protoplasmic stains. There is probably . 
going on in several of the layers under the epidermis, 
would seem to be more active close to the epidermis pears 
deeper layers. Three or four layers of cells immediately . 
the epidermis are of the same prismatic shape as hee 
cells proper, and almost as broad as high. Deepet -_ oe 
elongated in a direction perpendicular to the pie 
their walls become thinner. In vertical section they are te 
lie in definite rows. The distinction made by wes hold 
between outer and inner cortex in Lessonta ovala seems 10 
good also for Lessonia littoralis, but there is a gradual 
