Chapter II deals with the principles of a natural arrangement of the species within the genus, and in. Chapter III the importance of the different forms of hairs is considered from both the taxonomic and ecological standpoints. Chapter IV deals with seedling forms, those of six New Zealand members of the genus being described at length. The occurrence of long, many-celled, thin-walled hairs in certain early seedlings, but which are absent in the adults, is noted as requiring, special investigation.

Chapter V deals with the plant-geography of the genus The question as to whether it is primarily subantarctic or the contrary is discussed, and, on the assumption that certain characters are older than others, a plea is established for a northern rather than a southern origin But, on the other hand, it is shown how the present extensive occurrence of species not only in subantarctic South America, but throughout the subantarctic zone as a whole, together with New Zealand and parts of Australia, supports the theory of a southern origin.

Certain cases of discontinuous distribution are cited, especially that of the section Acrobyssinoideae, one species of which A. tasmonica Bitter, is endemic in Tasmania, but all the other species occur in the Chilian Andes from lat 39·50 south to Cape Horn and South Georgia. A. californico Bitter, endemic in California, belongs to a section not otherwise extending north of central Chile and Patagonia.

The occurrence of local forms is much commoner than has been thought hitherto, especially on the isolated subantarctic islands. This discounts considerably the general view that the supposed wide area of certain species was due to carriage of the fruits with their hooked barbs by birds. At the same time, the author considers the occurrence of closely related species in distant localities due to bird-carriage, and cites a Sandwich Island species, with its close relatives in south and central Chile, and the Acrobyssinoideae of the Magellan region and Tasmania, but absent in New Zealand. But such distribution was not frequent enough to hinder the evolution of many local forms.

It is a remarkable fact that the New Zealand species of Acaena possess the following peculiarities not present in members of the genus elsewhere: (1.) A stunted habit of growth, although their foreign allies are robust and large-leaved—as, e.g., the New Zealand forms of the polymorphic A. Sanquisorbae Vahl as compared with the Australian, and the three new species formerly referred to A adscendens Vahl in New Zealand in comparison with the true species of that name and A. laevigata (Ait.) Bitter of South America Nor can the alpine climate be responsible for the marked exhibition of this growth-form in the endemic New Zealand section Microphyllae. since the Chilian Andean species show no such character. (2.) The brownish colour of the leaves in certain varieties of the Sanguisorbae group and in A. microphylla Hook f, and such colouring is present in members of other families of New Zealand alpine plants Grey-coloured leaves, arising either through a covering of wax or through air-spaces between the cells, is another characteristic, but there are Chilian examples of similar phenomena

It would serve no purpose to draw up a synopsis of the author's arrangement of the New Zealand species A really critical study of his work is demanded on this point. Here only some general conclusions are noted regarding the limitations of species, important alterations, and so on.

First, it must be pointed out that a most comprehensive view is taken as to the species themselves, so that the polymorphic species are groups which do not exist at all as true entities. These species are subdivided into smaller but still comprehensive groups as subspecies, these into varieties which are the true entities of the flora, and these occasionally into forms.

A subspecies of A. avina A. Cunnn. is created called nanella Bitter, based on specimens sent by Cockayne to Berlin as introduced and collected on the Canterbury Plains, the author considering the stunted habit of specific importance.

Acaeno adscendens Vahl, hitherto supposed to extend to New Zealand, is restricted to plants of the Magellan region and Kerguelenland, while the New Zealand forms referred to A. adscendens are treated by the author as three distinct species—A. saccaticupula Bitter, A. hirsutula Bitter, and A. fissistipula Bitter, the two latter being closely related, but the former coming into a different subsection allied to that containing A. adscenden [ unclear: ] . Varieties are also described of all three species.

Acaena Sanguisorbae Vahl is restricted to Australian, Tasmanian, and New Zealand plants, the Tristan d'Acunha plant being referred to A. sarmentosa Car-michael and considered as restricted to that group, while the New Amsterdam plant is referred to A. insularis Citerne