Discusses the ranges of alpine species in U. S. and considers the possible migration routes of such species from Europe.

Lists those U. S. genera which he considers protean and describes the U. S. character of some genera which are protean in Europe.

Describes how he distinguishes introduced and aboriginal stocks of the same species.

Discusses difficulties involved in deciding which genera are protean in the light of some comments by H. C. Watson.

Comments on species with disjoined ranges; does not feel, despite CD’s expectations, that they tend to belong to small families.

Gives the proportion of U. S. trees in which the sexes are separate [see Natural selection, p. 62].

Believes, with CD, that extinction may be an important factor in explaining plant distributions, but sees no reason why the several species of a genus must ever have had a common or continuous area. "Convince me of that, or show me any good grounds for it … and I think you would carry me a good way with you". It is just such people as AG that CD has to satisfy and convince.

Feels that the crossing of individuals is important in repressing variation and perhaps in perpetuating the species, but instances some plants in which it cannot, apparently, take place.

States he has "misgivings about the definiteness of species". Believes there is some inherent tendency for plants to originate varieties. Cross-fertilisation is likely in most cases but sees difficulties with plants like Adlumia.