24 TRANS. ST. LOUIS ACAD. SCIENCE. 



The color of the leaves varies from deep or fresh green through 

 dull green to light glaucous. 



Of great interest and diagnostic importance is the edge of 

 the leaf. In some species (e.g. T. aloifolia, T. brevifolia, T. 

 rupicola) it is rough, or, as it is usullay termed, serrulate, and re- 

 mains unaltered through life. The teeth consist of small, irregular, 

 isolated cartilaginous knobs, each consisting of quite a large num- 

 ber of colorless prismatic or clavate cells, arranged in fan-shaped 

 or straight bundles. These are the "serrulate" or rough-edged 

 Yuccas. 



Others have "smooth-edged" leaves, (V. gloriosa, T. Trecu- 

 lianct) ; the edge, at first green, and often roughened with very 

 delicate and deciduous asperities, soon becomes discolored and 

 brittle, and in old leaves is apt to crumble off, or sometimes to de- 

 tach itself in a few short fibres, thus approaching the next form. 



The " filamentose," or fibrous-edged Yuccas ( T. fila?nentosa, 

 T. angustifolia, T. baccatd) constitute the third class. In these 

 the fibrous system of the leaves is much stronger and tougher 

 than in the last, and, the parenchymatous tissue soon withering 

 on the edge, the marcescent marginal fibres detach themselves as 

 more or less numerous, delicate, or coarse, straight, or often curled 

 threads, of a whitish, ashy or reddish color. In the young leaf 

 they are most conspicuous, especially near the involute point of the 

 leaf, but in old ones they sometimes become obsolete.* 



Some importance has been attached to the number of leaves, 

 which in healthy plants precede the development of the inflor- 

 escence, and there really is a relative difference in this respect in 

 different species ; but specific characters could hardly be based on 

 a condition which depends so much on external influences of soil, 

 climate, etc. From Dr. Mellichamp's notes it is evident that wild 

 plants, in good health, exhibit a great many more leaves than cul- 

 tivated ones, and that the number not rarely rises above one hun- 

 dred on one axis. 



The diagnostic characters derived from the leaves must be 

 adopted only with great .circumspection. The characters of the 

 edges of the leaves are the most constant and reliable ones, though 

 the abundance, thickness, and, still more, the length of the fibres, 



* A fourth form of leaves is described in the books as having- marginal spines, and T. 

 spinosa, H. B. K. nov. gen. I. 289, from Mexico, is quoted as the representative of this type. 

 But the inspection of the Berlin Herbarium proves this to be a factitious species, made up of 

 Yucca flowers (similar to those of Y. Treculiana) and the spiny leaves of Dasylirion 

 acrotriche. 



