69 
not only do we find them at times destitute of starch, but they are, 
without doubt, botanically identical with Smilax officinalis, H. e¢ B. 
I am aware that the form of the cells of the nucleus-sheath of the 
roots has been considered by physiologists as furnishing good marks 
of distinction between ‘the sarsaparillas of Central America and those 
of South America; and Schleiden declares that he can readily distin- 
guish them microscopically. But this theory does not seem to rest 
on any safe foundation. Let Mr. Bentley narrate what he saw :— 
“ When examined by the microscope the cells of the inner cortical 
layers are found to contain bundles of acicular raphides, and a large 
number of the starch granules already described. The pith also is 
found commonly to contain a number of similar starch granules. The 
breadth of the pith is usually from one to one and a-half, or perhaps 
a little more, that of the woody zone. In this respect it resembles the 
Honduras sarsaparilla, and, according to Schleiden, also all those sorts 
of sarsaparilla which are obtained from Central America. The cells 
of the liber, or, as it is called by Schleiden, the nucleus sheath, are 
elongated radially or from within outwards, and have walls which are 
thicker on the inner than the outer side. In this respect it resembles 
the South American and Mexican sarsaparillas, which Schleiden says 
always present this peculiarity. This microscopical appearance is 
remarkable, because, according to Schleiden, the Honduras and all 
the Central American sorts of sarsaparilla are characterized by having 
the cells of the nucleus sheath either square or somewhat elongated 
transversely, and all their walls of nearly equal thickness, and he 
believes that he can distinguish Central American from South Ame- 
rican and Mexican sarsaparillas from the appearances thus presented, 
combined with the different relative proportions of the woody layer 
and the pith already alluded to. But if this be true generally (which, 
so far as my experience goes, is not absolutely the case, having 
observed some sorts of Brazilian sarsaparilla in which the cells of the 
nucleus sheath were elongated somewhat in a direction from within 
outwards, and so far therefore agreeing generally with the anatomy of 
South American sarsaparillas according to Schleiden, but yet had 
their outer and inner walls of nearly equal thickness, and thus agree- 
ing with the Honduras variety), it is certainly not true in the present 
sort, for here we have a Central American sarsaparilla which agrees 
with Schleiden’s arrangement generally as regards the relative pro- 
portions of pith and woody layer, but differs from it in the cells of the 
nucleus sheath being elongated from within outwards or radially, and 
having walls which are thicker on the inner than the outer side.” 
