This comparatively wide area must have had, at some remote period, a steppe or semi-desert climate, under whose influences a xerophytic and partly calciphile flora developed and flourished, and it is likely that what we now have represents only a portion of this flora, many species having probably died out altogether.

This area must have been partly a peneplain (upon which alone R. paucifolius, it would seem, could “originate” and flourish), and would probably be conterminous with a range or ranges of hills with limestone rocks exposed and weathering into dust exactly as they now do on the small area here under observation. But such peneplain need not have consisted entirely of Tertiary limestone beds.

The area would be in the nature of a strip or belt, of no very great width and probably much interrupted, corresponding roughly to the shoreline or lines of the hypothetic Tertiary sea or seas. It would be conterminous with and more or less alternated with an area or areas of pre-Tertiary formation, probably lying to the north and east, as posited, e.g., by Cockayne (1911, pp. 343–44), by way of which probably the mesophytic flora would return when a more humid climate should prevail in this area. Upon this pre-Tertiary area the related species, R. chordorhizos, &c., would have originated and flourished, or that single species or form from which they and R. paucifolius trace their common descent.

The greater part of these limestone beds was destroyed by erosion of various kinds in subsequent ages, leaving only the present small isolated remnants, of which the Trelissick Basin is one of the largest.

It is impossible that by the elevation of the land 3,000 ft. or 4,000 ft. (Haast, Hutton, Park), and the consequent refrigeration and glaciation, the whole flora of the district (as has been thought) was driven to another tract, now non-existent, and returned with the subsidence of the land and consequent change of climate. “Return” of a calciphile flora over areas upon which the Tertiary beds had becn destroyed would be impossible, especially since, as we have seen, this flora as a unit is not a “traveller”; and we cannot escape the conclusion that this plant community has been represented within the area of the small basin, since it first established itself or “originated” in that neighbourhood.

Glaciation bears upon the question in two ways:—

(1.) Hutton (1900, p. 176), followed by Cockayne, correlated the supposed drought epoch, of which our flora shows signs, with the glacial epoch, which he placed in the older Pliocene period. This view was adopted by Cockayne (1901, pp. 280 et seq.); but that authority believed that at the height of the glaciation the eastern mountains (within which this area is included) might still support a xerophyte flora like that of the shingle-slips of the present day (Cockayne, 1911, pp. 348 et seq.).

The view of Speight (1911) and others is that the last glacial epoch is much more recent, that the drought period was correlated with it (Cockayne, 1911, p. 344), that the Tertiary deposits were continuous over a much larger area than is the case now (Speight, 1915, p. 54), that the Castle Hill area probably escaped glaciation altogether (Speight, 1917, pp. 323 et seq.), and that the Trelissick Basin at the height of glaciation was “probably a snowfield” (Speight, 1917, p. 323).

It would seem certain that a steppe climate or period of drought must have obtained here over a large area at least once (probably more than once) since Tertiary times, but to the present writer it seems quite uncertain whether this was coeval with and resultant from the glacial epoch or not

of direct adaptation or self-adaptation, as understood by Warming (1909) —it would seem probable that a single ancestral form of Ranunculus developed under conditions of extreme drought into a typical xerophyte, and that, after the conditions to which it had adapted itself had been modified or completely changed, this plant maintained itself against the competition of a mesophyte flora in certain localities—i.e., shingle-slips—in which it had an advantage, and in course of a long period of time, existing only in isolated areas completely separated from one another (one of which is the limestone area here described), it developed those comparatively trivial distinctive characters (especially in the cutting of the leaf) which now distinguish the “species” from one another.

According to De Vries (1912), however, such speculations and conjectures as to the conditions under which a species originated are idle, and can achieve no result. Speaking of “beautiful adaptations” to local conditions, he says: “In no case is it possible to tell whether the species have acquired these during their migration or during their stay in the new environment, or perhaps previous to their being subjected to the influence in question” (p. 592). Again: “Adaptations to new conditions [which are conceded] depend upon characters which were inherent in the species before it arrived in the new environment. The characters themselves are not the effect of the external influences considered” (p. 579). Such characters, it is contended, cannot be good specific marks; they fall within the range of “fluctuations” (as distinguished from mutations) and “cannot lead to constant races” (p. 540). The species thus modified or adapted remains essentially the same, and will, if replaced in the favourable conditions, resume its older form (as in the classic experiments of Cockayne upon seedling forms, and those of Bonnier upon alpine plants). The sole condition required in the plant is therefore “high plasticity.” We must not say that a species originated under the stimulus of its environment, or that it acquired new characters in response to changed conditions: that would be confusing cause and effect. “Fitness for present life-conditions… can hardly be considered as a result of adaptation, and we have to recur to previous hypothetical environments to explain the much-admired adjustments. All speculations of this kind are merely reduced to more or less plausible and more or less poetical^* considerations” (p. 574). It is concluded that “geological changes of climate may have been accompanied by the production of new forms, but there is no evidence that this has occurred in such a way as to provoke directly useful changes”; that “the characters of local and endemic types do not betray any definite relation to their special environment”; and, finally, that “the facts which are at present available plead against the hypothesis of a direct adjusting influence of environment upon plants, and comply with the proposition of changes brought about by other causes and afterward subjected to natural selection” (p. 595). The author then restates his personal belief “that the species-making changes occur by leaps and bounds, however small.”

If these conclusions be accepted, the case of Ranunculus paucifolius and its associates may be thus considered in their light. It is generally accepted that a period of more or less severe drought or “steppe climate” has been passed through by a great part at least of the flora of New

Zealand. The particular community here studied shows this with especial clearness, consisting as it does of a small association of plants all of which show very definite xerophytic adaptations, while some of them can exist only under certain very special and peculiar edaphic conditions such as may have obtained more widely in the past. The conditions governing plant-life before and during this period of drought may be supposed to have been much the same as those of the Sahara at the present time, thus described by De Vries (after Battandier): “Originally this region must have had an ordinary degree of rainfall and moisture…. Then… the rainfall must have slowly diminished, taking centuries… to reach the conditions which now prevail. The consequent changes in this flora must have been correspondingly slow, and must have consisted mainly in the disappearing of the larger part of the species; first of those which were dependent on the higher degree of moisture; then of others; until at the present time only the most drought-resisting forms are spared” (pp. 589–90). He proceeds to show that no specific changes, probably, were brought about by this process; that a large number of the species of this arid region are monotypic genera, each genus consisting of a single species; whereas, “if there had been any degree of adaptation during this whole period of increasing dryness, new species would have been produced —from those forms which by their own inherent capacities would be the very last to be threatened with extermination. These genera would therefore have produced quite a number of smaller or even of larger species, adapting themselves more and more to the changing conditions and stocking the desert, in the same way as other deserts have been stocked, from adjoining countries!” As this has not happened, it is concluded “that the single species… have not undergone any change in the direction of drought-resistance, but have simply been those which happened to be the best fitted for the life in the desert. A thick epidermis, a small display of leaves, long and deep roots, were the main qualifications for this choice” (p. 590).

Then, in our case, we assume that the moister climate re-established itself; the mesophyte flora which had been destroyed here, but had maintained itself in some adjoining land where the conditions remained favourable, returned and gradually repeopled the desert or semi-desert, while the xerophytes retreated before it to those places, such as shingle-slips and areas like the small hollow at Castle Hill, where they had an advantage and have subsequently maintained themselves. But, in contradistinction to what has been said above, we must accept the following propositions as to this community of plants:—

(1.) The species here studied—e.g., Ranunculus paucifolius, Lepidium sisymbrioides, Oreomyrrhis andicola var. rigida, and Poa acicularifolia— all existed and held their own among the pre-drought mesophyte flora, but not perhaps exactly in their present form, since “adaptations” are not denied except as differential marks of new species. The only alternative is that they have originated, some or all of them, since the period of “drought.”

(2.) All these species must have had a high degree of plasticity, and thus they are able gradually to accommodate themselves to the increasingly severe drought; but all must have had already, at the beginning of the period, a definitely drought-resisting structure, and this was not at any time acquired by any of them in response to any external stress, and it would be at first quite useless to them.

(3.) Each of these species originated by a mutation or sudden change involving the introduction of at least one quite new unit-character,^* and this must have happened at some period anterior to that of the drought, not as a result of any such condition. Like other differential characters in general, those induced by this mutation would be at first perfectly useless (De Vries, 1912, p. 534), and the changed form would get its advantage only by the chance of the occurrence of the drought. The new character or characters then became useful; but we must resist the temptation to regard the useful character (e.g., the excessively thick and coriaceous leaf or long thick roots of the xerophytic Ranunculi) as an adaptation to the needs of the new external condition.

(4.) As “adaptations” can in this case not be denied, it follows that all the changes which are truly “adaptations” in these species are of the nature of “fluctuations,” and if any of them be cultivated under more favourable conditions the “adaptations” will disappear; the plant will then retain only so much of its xerophytic character as it had at the beginning of the drought, which gave it its initial advantage over others, and which was the result of some previous mutation. Until each plant of the community, therefore, has been so transplanted and tested it is impossible for us to tell which of its characters ought, and which ought not, to be regarded as differential specific characters; and it follows that the status of each is doubtful except where the plant has no near relatives at all among existing plants.

(5.) It is very improbable that the species of this community were all produced in the early stages of the drought by mutation. It is assumed “that the origin of new forms is not due to a hard struggle, but is promoted by a luxuriant environment and by easy conditions of development” (De Vries, 1912, p. 520). It is shown that a species (or genus) which is in a “state of mutability” may produce whole groups of new forms, even “swarms” (as in the case of Draba or Viola in Europe), though sometimes apparently such changes are only sporadic (p. 549). In this case it must be supposed that at some more or less remote period before the drought each of the genera Ranunculus, Lepidium, Oreomyrrhis, Myosotis, and Poa passed through a “mutation period” and threw off numbers of new species, some of which would immediately perish, while others would maintain themselves for shorter or longer periods under the stress of natural selection, and finally the species here perpetuated would alone survive under the fierce stress of the drought until rejoined by their relatives under the new climatic conditions.

(6.) Narrowing down the proposition to the particular genus and species here studied, we must believe that there existed at the beginning of the period of drought a species (or possibly more than one) of Ranunculus which had originated by mutation at some period (as to which it is useless to speculate) having peculiarly thick leaves, long roots, and other characters which gave it an advantage when the drought began to be severe. These characters, however, had been acquired by it owing to causes which are completely dark to us, not in response to any external stimulus or stress of environment. Fortunate in possessing these characters, it continued to live when other less-favoured Ranunculi perished, and it may or may not have changed under the new conditions, adapting itself thereto. But if it did so change it acquired no new unit-character; and all its modification remained mere “fluctuations,” and under more favourable conditions