256 Gates.—A Systematic Analytical Study of 
fruits ; also in certain minor characters of foliage and pubescence. But the 
two species run into each other and probably intercross. 
D. Smithii and D. Hookeri form a pair differing in having respectively, 
(1) larger whitish and smaller greenish flowers, (2), densely short-hairy or 
glabrous style and ovary, (3) leaves mostly subcordate or cordate at base; 
(4) in D. Smithii the leaf margins are ciliate. D. Hookeri differs from all 
the others in having flowers which are broad and truncate at base. D. trachy- 
andrum probably originated from D. Hookeri through a mutation. It 
differs strikingly in having hairy anthers, but the leaves are also less deeply 
cordate and distinctly broader towards the apex, and the ovary is glabrous. 
Thus it will be seen that for the most part the species of this genus are 
sharply separated from each other by ‘ presence and absence ’ characters, 
and only to a minor extent by quantitative characters. The latter are 
likely to be often merely by-products, as it were, of the more striking 
germinal changes. It is impossible to suppose that, e. g., stages in the 
development of the reticulation of the fruit, the lobing of the stigma, or the 
hairiness of the anthers could be of use to the plant. Indeed, it seems 
fairly clear that such differences are the result of sudden ‘ chance variations ’ 
and have no significance in the economy of the plant’s life. Having 
appeared, heredity perpetuates them, and so new varieties or species come 
to be established. Their comparative recency of origin can perhaps be 
judged to some extent by the relative areas they now occupy. 1 We should 
then expect D. maculatum and D. trachyandrum to have originated rela¬ 
tively recently, while D. trachycarpum , which occupies a much wider area, 
may have originated from D. lanuginosum at an earlier time. There are, 
however, difficulties in applying this view in detail even to such a relatively 
simple genus as Disporum . But the underlying idea of definite variations 
arising in certain localities and afterwards spreading, rather than the hypo¬ 
thetical accumulation of small differences and the subsequent elimination of 
intermediates, corresponds with the facts of such a genus as we observe 
them. In the light of the facts of mutation, single definite steps, or in some 
cases a few steps, from species to species are by no means improbable, and 
it appears that all the conditions in the North American Disporums can be 
explained by their aid, without resorting to the accumulation hypothesis. 
The genus Uvularia is confined to eastern North America. It contains 
two species, U. perfoliata and U . grandiflora , and a third, U. flava , which 
is known only from a figure and description in Smith’s ‘ Exotic Botany 
and is probably a hybrid between the other two. Both these species range 
from Quebec nearly to Florida, but U. grandiflora extends farther west into 
1 This was originally written before the interesting papers of Willis appeared, applying 
this conception to the floras of Ceylon and New Zealand. 
See Willis, J. C. : The Evolution of Species in Ceylon, with reference to the Dying Out of 
Species, Annals of Botany, vol. xxx, 1916 ; The Distribution of Species in New Zealand, ibid., 
pp. 437-57 ; and several other recent papers. 
