430 Farmer and Freeinctn. — On the Structure and 
stachys. It cannot, however, be referred to a definite cell-band 
which can in any way be identified as ‘ the innermost layer 
of the cortex,’ for the characteristic markings sometimes 
extend over two bands of cells, and at other times they extend 
also to cells which obviously appertain to still more external 
cortical layers. Thus, in this plant at any rate the endo- 
dermis, whilst retaining its physiological importance, lacks 
the definiteness which alone can give it that degree of morpho- 
logical importance which is ascribed to it by many investi- 
gators. The cells composing the endodermis are strongly 
cuticularised over their radial and often over part of their 
tangential walls, a fact which is abundantly demonstrated 
after treatment with strong sulphuric acid. The pitting of 
the walls is also very characteristic, and gives a peculiar 
striped appearance to the cells. 
Within the endodermis a somewhat broad layer of paren- 
chymatous cells which are somewhat elongated in form can 
be readily seen, and this forms the pericycle of the stele. 
Immediately within this band the sieve-tubes are disposed. 
They form a very easily recognizable layer, both on account 
of their form and especially by reason of their thickened 
cellulose walls. In respect of their curious walls they resemble 
other Ophioglosseae as described by Poirault 1 ; and while 
rather exaggerating the peculiarities observed by him for 
Ophioglossum and Botrychium, they plainly belong to the 
same type. When seen in longitudinal section they are 
found to consist of elements perhaps about eight or ten times 
as long as their diameter, and of a somewhat irregular form. 
The reason for this latter peculiarity will appear subsequently. 
Their walls, are, as has been said, very thick, and this is due 
to the swelling of the middle layers of the walls in which 
pectic substances are abundantly present. The sieve-areas 
are freely distributed over the lateral walls, and they occur, 
though not so frequently, on the end walls also, where two 
elements are in contact. The sieve-areas are very well marked 
1 Poirault, Recherches anatomiques sur les Cryptogames Vasculaires, Ann. Sci. 
Nat. (Bot.), ser. VII, t. XVIII. 
