90 A. H. Graves, 
a. Epidermis. 
(1) Chloroplasts. 
An interesting point brought out by a study of the epidermis, 
and shown especially well in the living condition, is the fact that 
here most of the photosynthesis is carried on, for the cells contain 
large numbers of chloroplasts (Text-fig. 12). The discussion of the 
causes of this condition, being of an ecological nature, will be post- 
poned for the present. 
(2) Marginal Teeth. 
The teeth, mentioned above, p. 88 (Pl. VI, fig. 26; Text-fig. 11), 
have already been observed in the leaves of 
Ruppia maritima by Sauvageau (1894, II, p. 209), who 
says, ‘‘a son sommet [1. e. of the blade] la plupart 
des cellules terminales se prolongent en dents 
arrondies composees de 1—2-3 cellules. Des dents 
semblables, mais plus espacées, se retrouvent sur 
les bords lateraux du limbe et font legerement 
saillie.” Ina young stage, as in Pl. VII, fig. 39, the 
teeth appear mostly at the end of the leaf primor- 

Figure 12.— Set : 
Portion of epi- dium; later they may be found down the margins 
dermis of leaf, of the leaf anywhere from 4 to } of the whole 
showing chloro- distance from the apex. Finally, in the adult leaf 
plasts. Drawn 
from living leaf. they rarely extend more than 7 of the distance 
>< 330. from the apex. 
(3) Secretion Cells. 
In the epidermis I have observed also the cells referred to by 
Sauvageau (1891, II, p. 209) as “cellules sécrétrices,” containing, in 
alcoholic material, a granular or a homogeneous content of a 
brownish color, and somewhat larger than the ordinary epidermal 
cells (Pl. VI, fig. 30; Pl. IH, fig. 8). They appear to be identical in 
nature with those of the stem. Although Sauvageau asserts that they 
are most abundant at the edges and at the apex of the leaf, yet 
they appear to me often to increase in numbers towards its base, 
including that part which adjoins the sheaths, as well as in the 
sheaths themselves (Pl. VI, fig. 25). According to Sauvageau 
these peculiar cells occur also in Posidonia, Cymodocea and Halo- 
dule. He has ascertained the presence in them of tannin, probably 
in combination. In my permanent slides their contents were almost 
invariably granular, assuming, with the triple stain, either a crimson 
or a brilliantly refracting yellow color—more often the latter. 
With ferric chloride the contents stained a brownish black and the 
same result was given on treatment with potassium bichromate. 
