624 Willis .— The Origin of Species by Large , rather than by 
great importance, such as the number of the stamens, which varies from six 
to a very large number. Rumination has no use-value as against non¬ 
rumination, and there are no intermediates conceivable that would have any 
value. Either the same character has been acquired over and over again, or 
genera are not units of descent, but are polyphyletic, as has often been 
suggested (cf. 11 , p. 446; 20; or 1 , p. 119). 1 This latter supposition, however, 
does not seem to cover more than a very small proportion of the cases, and 
we are driven to the conclusion that the same indifferent, but systematically 
very important, character, which rarely admits of intermediate stages, and 
which therefore cannot have been acquired gradually or by selection, has 
been acquired over and over again. Nothing but sudden acquisition, or in 
other words ‘ large ’ mutation, seems capable of explaining this phenomenon, 
which is familiar to every worker in taxonomy. 
But the acquisition of ruminate endosperm is not even a question 
of acquisition by one genus only and not by another. In the four following 
palms part of the genus shows it and part does not, so that it becomes 
merely a specific character, and it is clear that one species may acquire it 
and another not: 
Ruminate. 
Euterpe oleracea 
Oenocarpus , § II 
Hydriastele , p.p. 
Nenga, §§ I, II, III 
Non-ruminate. 
Euterpe , other spp. 
Oenocarpus , § III 
Hydriastele , p.p. 
Nenga, §§ IV, V 
Or one may take evidence of a different kind by taking a single family 
and considering its range of variation. Take for example the Rubiaceae. 
None of the important characters of the family run unchanged throughout 
it, and one finds, for instance (the contrasted character is the usual one in the 
family): 
alternate leaves in Didyniochlamys ; 
whorled leaves in Fadogia ; 
pinnate leaves in Pentagonia ; 
gland-dotted leaves in Rustia ; 
intrapetiolar stipules in many ; 
leafy stipules in Galieae ; 
dioecious flowers in Kotchubaea ; 
zygomorphic flowers in Capirona ; 
solitary axillary flowers in many ; 
male and female inflorescences often very different ; 
male and female flowers so different in Melanopsidium that they were 
formerly described as different genera ; 
flowers united in pairs in Morinda , &c. 
male flower 4-5-merous, female 8-merous, in Thieleodoxa ; 
1 Sir N. Yermoloff (20), in his studies of the Diatomaceous genus Naviada , appeals to a principle 
of integration rather than to the differentiation of types, i. e. to the building up rather than to the 
breaking down of types. 
