Art. XXXIII.—On the Burning and Reproduction of Subalpine Scrub and its Associated Plants; with Special. Reference to Arthur's Pass District.

Transactions and Proceedings of the Royal Society of New Zealand, Volume 31, 1898, Page 398

Art. XXXIII.—On the Burning and Reproduction of Subalpine Scrub and its Associated Plants; with Special. Reference to Arthur's Pass District. By L. Cockayne. [Read before the Philosophical Institute-of Canterbury, 4th May, 1898.] Plates XXXV.–XXXVII. Up to the present time no scientific account of the effect of fire on New Zealand vegetation, illustrated by accurate observations, has been published. From time to time generalised statements, founded on slight or insufficient data, have appeared in the “Transactions of the New Zealand Institute “and elsewhere. Thus, for instance, the Rev. P. Walsh writes*“On the Disappearance of the New Zealand Bush,” by the Rev. P. Walsh (Trans. N.Z. Inst., vol. xxix., p. 496).: “Nature makes a brave effort to reclothe the hills and gullies of New Zealand in her verdant mantle, and if let alone would bring her work to completion. Under. favourable circumstances seedling trees soon make their appearance, and if

protected from injury would in due time attain to maturity. Of course, anything like a real restoration of the original; bush is out of the question, but the second growth has a beauty of its own, which is by no means to be despised…. It is interesting to notice that everywhere the trees which are characteristic of the locality are not long in making their appearance.” The late Mr. T. Kirk, F.L.S., has gone into the matter with somewhat greater detail.*“The Displacement of Species in New Zealand,” by T. Kirk, F.L.S. (Trans. N.Z. Inst., vol. xxviii., p. 16). “On the west coast of the South Island,” he writes, “much of the New Zealand forest, when burnt off, is temporarily replaced by a robust growth of a large native groundsel (Erechtites prenanthoides, DC.), which often attains the height of 5 ft., most of it disappearing before the close of the third year, when its place is taken by fern, or more rarely by shrubs and trees”; and, regarding a fire in the Hope Valley, he states: “The burnt area on each side of the road-line was thickly dotted with the rare pine, Podocarpus acutifolius, T. Kirk, although very few specimens of the plant were to be seen in the immediate vicinity”; and he concludes his remarks thus: Much, however, has yet to be learned with regard to phenomena of this kind in “New Zealand.” During a stay of six weeks' duration this summer—December, 1897, and January, 1898—on the summit of Arthur's Pass, I was enabled to take the fairly copious notes embodied in this paper on the effect of two fires which had devastated the vegetation of that locality, the one a recent and the other, a burning of more distant date. Such notes will, I think, tend to show what would be the ultimate results of fires in localities similar to Arthur's Pass, to such, indeed, on the western side of the dividing-range, or on the eastern within the limits of the heavy western rainfall and its accompanying misty weather. Regarding the first fire, I can give no exact date; possibly it took-place twenty or more years ago, or there may have been several fires during the period that the pass has been used for traffic. But of the recent fire more exact information is to hand: it took place in the year 1890, and was the work of the Midland. Railway survey. This fire, originating somewhere on the left bank of Peg Leg Creek, hear its junction with the Otira River, crossed that river and ascended the Westland spur of Mount Rolleston to a height of 1,000 m. or more, while on the Arthur's Pass side it followed the Otira River for about 1 kilom. towards its source; also, spreading round the Canterbury spur of Mount Rolleston, it ascended

that mountain almost to the limit of the shrubs at 1,200 m., finally burning the whole of the shrubby vegetation on the more level portions of the pass—in places crossing the road—except, a few patches here and there, as far as the Fagus forest in its south-west corner, the outskirts of which were in places destroyed. It is also quite possible that this same fire was responsible for the burning of most of the vegetation on the Canterbury side of Hill's Peak. Previous to the last fire or fires the whole of Arthur's Pass except swamps and ground liable to flood was covered by more or less dense subalpine scrub. This subalpine scrub, Or dense mass of rigid branching shrubs, is one of the most noteworthy features of New Zealand vegetation, and has always been made a subject of considerable comment by New Zealand explorers and the like. The Rev. W. S. Green thus wrote of it:*“The High Alps of New Zealand,” by the Rev. W. S. Green. London, 1883. “There were a number of other bushes with strong gnarled stems and small leaves”; these “combined to form as ungetthroughable an obstacle as it was possible to imagine.” Haast also speaks of such shrubs as “impenetrable scrub,” and, writing of Meins Knob, he says,†“Report of the Head Waters of the River Rakaia,” by Julius Haast, Ph.D. Christchurch, 1886 (page 20). “For botanical purposes I returned to the foot, of the hill through the bush, a herculean task, particularly for one of portly dimensions, as we had often to lie flat on the ground and crawl through or walk over the tops of the branches.” T. Kirk describes‡” On the Flowering-plants of Stewart Island,” by T. Kirk, F.L.S. (Trans. N.Z. Inst., vol. xvii., p. 220). graphically a similar scrub on Mount Anglem, Stewart Island; and Dr. Diels, in his recent work on the biology of New Zealand plants, §“Vegetations-Biologie von Neu-Seeland,” von L. Diels. Leipzig, 1896 (pp. 261–263). treats of this subalpine scrub at some length. The subalpine scrub occurs usually just above the forest-line, at first intermixed with the forest-trees and afterwards forming a distinct belt and barrier between the forest and the grass-line for a varying distance, and ending usually at 1,000m. to 1,200 m., or even higher. Such scrub occurs more or less on all the high mountains; but on the dryer ones—such as those of Central Otago, East Nelson, Marlborough, or the eastern portions of the Southern Alps—not forming a distinct belt, but only patches in places. Its tendency to burn is well exemplified by the local name, “turpentine scrub.” Growing in association with this scrub, in places where it is not too dense, and especially towards its highest altitudinal limit, are the most striking herbaceous plants

of our flora. The scrub itself, top, consists in the mail of beautiful flowering-shrubs, a great number of which are how favourite garden plants in many parts of the world. Should this subalpine scrub be set on fire during a dry season very large areas will be burnt to the ground, and, except in some few protected spots, apparently completely destroyed, together with the contained herbaceous plants and grasses; When it is borne in mind that musterers light fires promiscuously on the mountains to show their whereabouts to their fellow workers, it may readily be seen that, unless nature does something towards its replacement, this most interesting feature of the New Zealand flora bids fair to become a thing of the past. That nature does not replace this loss is at the present time the commonly received opinion. Mr. A. Harper, in his work on the Westland Alps, makes the deliberate statement*“Pioneer Work in the Alps of Now Zealand,” by Arthur P. Harper, B.A. London, 1896. that “the scrub never grows again when burnt,” and consequently he set fire to various patches, so as to provide an easy route up the mountains in time to come. That this statement is not in accordance with facts will be seen from the results of my investigations, as stated at the, close of this paper. The flora of Arthur's Pass and its vicinity may be naturally divided into two sections—-eastern and western—the differences between which can be at once perceived even by the untrained eye, the eastern being more lowly in growth than the western, and having as near neighbour a Fagus forest, while the western is much taller, greener, and more luxuriant. Here Fagus is entirely absent, while Dracophyllum traversii, a most remarkable Epacrid, forms conspicuous clumps. This difference is one of habit and percentage of component parts rather than of great difference is species, the dominant shrubs of the western not always being the same as those of the eastern division. After making a general examination of. the whole, certain spots suggested themselves as eminently suitable for determining—first, the nature of the former vegetation; and, second, the plants which had appeared since the more or less complete destruction of the original scrub. These spots all contained patches of considerable size which the fires had completely spared, some belonging to the Midland Railway fire and others to that of earlier date. Bach spot or section, as I propose to call it, I examined carefully, making, with regard to the living scrub, a list of its species, noting their height, the quantity or proportion of each species, the seedlings growing under their shade, with the quantity and size of such

seedlings. With regard to the adjacent burnt portions, I made lists of seedling plants, their size, their number in a stated area, and their numbers with regard to each other; noted also those plants which were only killed to the ground, and had come up from the old stock, and those also which escaped the fire altogether. In one instance, also, I added, to my list all the introduced species not indigenous which had, made their appearance.* In most of the sections this was not necessary, for introduced plants take little or no place. Some attention was also given to variation of seedlings growing under the new conditions of light and moisture, but such I reserve for consideration elsewhere. The various sections, being mostly easy of access, are readily available for future observations, and the struggle for survival between the present small plants can easily be watched and recorded by any future observer. Section A comprises the country at the Canterbury end of the pass in the vicinity of the Pass Creek, on both sides of the West Coast Road. Near the Fagus forest, reaching to the edge of the creek, is a piece of unburat scrub, while another piece lies just over the main creek, on its northern side and close to the road. With regard to the burnt portion, there is a fine example for investigation between the two branches of the creek. The following list is drawn up from notes on these portions, and written on the spot, as were, indeed, all the notes from which this paper is made up :— Former Constituents of Section A (arranged roughly in order of greatest quantity). 1. Dracophyllum longifolium, Br.; covering half the area, 1.35 m. in height. 2. Phyllocladus alpinus, Hook. f.; forming nearly one-sixth of the whole. 3. Gaya lyallii, Jack, and Hook.; abundant, especially near the forest; 4 m. to 6 m. in height. 4. Panax colensoi, Hook. f.; 1.35 m. in height. 5. Coprosma cuneata, Hook. f. 6. ", Hook. f. Scattered or in clumps here and there, but in no great quantity. 7. Veronica odora. Hook. f. Scattered or in clumps here and there, but in no great quantity. 7A. ", †This Veronica is the most abundant form in Westland, to the exclusion of all others except V. salicifolia at a certain altitude. Mr. T. Kirk, to whom I referred a specimen, named it V. traversii, var. approaching V. lævis. I think, however, it is as good a species as most of our Veronicas, varying to no great extent, and occupying wide areas. I therefore propose to call it provisionally “Veronica subalpina.” sp. nov. Scattered or in clumps here and there, but in no great quantity. 8. ", J. B. Armst. Scattered or in clumps here and there, but in no great quantity. Scattered or in clumps here and there, but in no great quantity. 9. Cassinia vauvilliersii, Hook, f. 10. Coprosma ramulosa, Petrie.

11. Coprosma serrulata, Hook. f. 12. Myrsine nummularia, Hook. f. 13. Senecio bidwillii, Hook. f. 14. Pittosporum rigidum, Hook. f. 15. Pittosporum var. with white flowers. 16. Dracophyllum uniflorum, Hook. f. 17. Aristotelia fruticosa, Hook. f. 18. Coprosma,* This may be a ciliated var. of Coprosma parviflora, Hook, f., as indicated by Cheeseman, or it may be quite possibly an undescribed species. sp. with ciliated leaves. Beneath the dense portion of the above scrub grew very few herbaceous plants and grasses, but, where thinnest, Aciphylla colensoi, Hook, f.; Phormium cookianum. Le Jolis; Erechtites glabrescens, T. Kirk, Hierochloe alpina, Rœm. and Schl.; Danthonia raoulii, Steud.; various Epilobiums; and possibly many other plants either not growing under the present living scrub or unnoticed by me (Plate XXXV.). Ferns were represented by Lomaria alpina, Spreng.; Lomaria procera, Spreng.; and Hypolepis millefolium, Hook. Where the height of the scrub is not stated above, the average would be perhaps from 1 m. to 1.5 m. Seedlings Under Unburnt Portion of Section A. (1.) Under Phyllocladus. Aristotelia fruticosa, Hook. f.; 0.07 m. in height. Coprosma parviflora, Hook. f. Olearia nummularifolia, Hook, f. Pittosporum rigidum, Hook, f.; quite small; only developed to its fourth leaf. All these were growing in considerable quantity in the decayed leaves and matted roots, which form a moist soil for a depth of 0.06 m. (2.) Under Podocarpus nivalis, Hook. f. Nil. (3.) Where Sunlight can partially penetrate. Veronica canterburiensis, J. B. Armst. Gaya lyallii, Hook, and Jack. Panax colensoi, Hook, f. (4.) Where the Scrub is dying out with Old Age. Panax colensoi, Hook. f. Veronica canterburiensis, J. B. Armst. Dracophyllum longifolium, Br. Veronica subalpina, mihi (ined.). Panax anomalum, Hook. f. Many of these seedlings are 0.33 m. high.

Section A.—Plants Which Have Appeared in the Burnt Portion Since the Fire. The locality specially examined lies between the two branches of the creek, and on Hill's Peak side of West Coast Road. (1.) Shrubs. Cassinia vauvilliersii, Hook, f.; in quantity. Veronica odora, Hook.; plentiful where ground is wet. Coprosma parviflora, Hook, f.; in large quantity. Gaya lyallii, Hook. and jack.; growing from old stump. Aristotelia fruticosa, Hook. f.; in quantity. Veronica canterburiensis, J. B. Armst.; in large quantity. Coprosma serrulata, Hook. f.; seedling plant (?). Coprosma acerosa, A. Cunn., var.; in quantity. Veronica subalpina, sp. nov.; in large quantity. Coprosma cuneata, Hook, f. Coprosma sp. with ciliated leaves. Gaultheria antipoda, Forst. Coprosma ramulosa, Petrie; seedling (?). (2.) Other Phanerogams. Epilobium nummularifolium, A. Cunn. Epilobium sp. allied evidently to E. glabellum, Forst. Epilobium pubens, A. Rich. Acæna sanguisorbæ, Vahl, var. Acæna glauca, sp. nov. (ined.). Phormium, cookianum, Le Jolis. Luzula campestris, DC., var. Festuca duriuscula, L., var. Ranunculus lyallii, Hook. f. Ranunculus plebeius, Br. Celmisia armstrongii, Petrie. Celmisia coriacea, Hook, f. Celmisia spectabilis, Hook. f. Celmisia longifolia, Cass., var. Celmisia flaccida, sp. nov. Hierochloe alpina, Rœm. and Schultz. Danthonia pilosa, Br. Geranium, microphyllum, Hook, f. Wahlenbergia saxicola, A. DC. Veronica lyallii, Hook. f. Euphrasia cockayniana, Petrie. Erechtites glabrescens, T. Kirk. Pratia angulata, Hook. f. Ourisia macrophylla, Hook. Ourisia macrocarpa, Hook, f. Ligusticum haastii, F. von Müell.

Ligusticum aromaticum, Banks and Sol., var. Gentiana bellidifolia, Hook. f. Viola filicaulis, Hook, f. Astelia nervosa, Banks and Sol. Aciphylla colensoi, Hook. f. Agrostis canina, L., var. Raoulia parkii, Buch. Angelica gingidium, Hook, f. Viola cunninghamii, Hook, f. Uncinia compacta, Br., var. Cotula perpusilla, Hook. f. Poa anceps, Forst., var. Uncinia, sp.; tall, with, narrow leaves. Galium, sp. Hydrocotyle novæ-zelandiæ, DC. Coriaria angustissima, Hook. f. Plantago brownii, Rafin. Mimulus radicans, Hook. f. Geum pusillum, Petrie (?). Senecio bellidioides, Hook. f. Schænus pauciflorus, Hook. f. Gnaphalium bellidioides, Hook. f. Senecio lyallii, Hook. f. Myosotis forsteri, Rœm. and Schultz. Introduced Plants. Stellaria media, Witt. Holcus lanatus, L. Trifolium repens, L. Trifolium pratense, L. Rumex acetosetla, L. Ferns. Hypolepis millefolium, Hook. Lomaria alpina, Sprengl. Aspidium aculeatum, Sw., var. To give some idea of the rapid reproduction of certain shrubs, in one square metre of ground were counted: Veronica subalpina, 118 plants, 0.34 m. to 0.02 m. in height; Veronica canterburiensis, 4 plants; Veronica odora, 1 plant; Cassinia vauvilliersii, 35 plants; Coprosma (with ciliated leaves), 2 plants: total, 160 plants. The plants enumerated were all growing among the remains of the dead standing shrubs, and the list gives an example of how a vegetation of herbaceous plants and grasses will occupy the ground in the first instance, to be afterwards choked put by a new growth of shrubs. To summarise the results, the old scrub—consisting principally of Gaya lyallii, Phyllocladus

alpinus, and Dracophyllum longifolium—will be succeeded by an equally thickly growing scrub, consisting principally of various Veronicas and Cassinia vauvilliersii, i.e., in this particular portion examined the former leading shrubs will be completely absent, except Gaya lyallii, represented by one solitary specimen reproduced from an old plant, while a new scrub will take its place, consisting mainly of species very scantily represented in the old. Section B: A very small section on the left-hand side of the “West Coast Road, nearly opposite the living portion of Section A, beyond the Pass Creek. This was before the fire a thicket of Veronica subalpina. It is now reproducing itself much as it was before. Young plants of 0.26 m. in height abound, growing amongst the remains of the burnt vegetation. Close by, some dead stumps mark the former presence of Phyllocladus alpinus, which, as usual, has not reproduced itself. This section being extremely wind-swept, the bushes only attain to a height of about 1 m. at most, so do not become dense enough to hinder a growth of herbaceous plants, which thrive well under the shelter, and where burnt have been quickly reproduced, their growth perhaps stimulated by the excess of potash. Section C comprises the portion of the pass between the burnt Fagus forest and the creek, special attention having been given to the portions near the forest. Here is only one living patch, a small piece of isolated Fagus cliffortioides, and, as that was not typical of the remainder, no notes were taken of its contents. Dead shrubs in abundance, and a few which escaped the fire, give good evidence of what the original scrub consisted, and which seems to have been almost the same as that described in Section A, with the exception, that Phyllocladus formed quite three-quarters of the whole of that portion nearest to the bush. Nearer to the creek Dracophyllum was more common. Section C.—Plants Which Have Appeared in the Burnt Portion Near The Bush Since the Fire. 1. Hypolepis millefolium, Hook. f.; 0.30 m. in height. This forms the most striking feature of the new growth, and owes its origin most likely to the proximity of this scrub to the forest, whose shelter would be conducive to an original fairly vigorous growth of fern. 2. Phormium cookianum, Le Jolis; in quantity. 3. Panax colensoi, Hook, f.; reproduced after being burnt to the ground; 0.51 m. in height. 4. Senecio elædgnifolius, Hook, f.; 0.25 m. to 0.20 m. in height; seedling plants in abundance.

5. Aristotelia fruticosa, Hook. f.; seedlings plentiful in patches; height, from 0.50 m. to 0.45 m. 6. Gaya lyallii, Jack. and Hook.; reproduced from burnt stump; height, 0.35 m. 7. Podocarpus nivalis. Hook. f.; a few plants; 0.18 m. in height. 8. Coprosma parviflora, Hook, f.; plentiful in places; height, 0.48 m. to 0.40 m. 9. Veronica canterburiensis, J. B. Armst.; seedlings; 0.36 m. to 0.37 m. in height; occurs in quantity. 10. Veronica subalpina, sp. nov.; seedlings; plentiful in places; 0.30 m. in height. 11. Cassinia vauvilliersii, Hook, f.; seedlings; in medium quantity; 0.26 m. to 0.32 m. in height. 12. Coprosma cuneata, Hook, f.; reproduced from burnt stump; scarce. 13. Veronica salicifolia, Forst.; only one plant observed; 0.48 m. in height. 14. Olearia illicifolia, Hook, f.; in small quantity;. seedlings; 0.14 m. in height. 15. Coprosma serrulata, Hook; f.; from an old plant; 0.50 m. in height. 16. Coprosma ramulosa, Petrie; in large patches; probably little damaged by fire, and may have formed considerable portion of original undergrowth. 17. Myrsine nummularia, Hook, f.; also, in quantity, and perhaps little damaged in first instance. 18. Phyllocladus alpinus, Hook, f.; seedlings very rare and very small. 19. Clematis australis, Kirk; most likely from an old plant. Of the new vegetation, Veronica canterburiensis will most likely, in time, form the greater part. It will be interesting to note the struggle for existence between the fern and the various seedlings. As the fern dies to the ground yearly, perhaps in the end it will succumb. I did not note many of the herbaceous plants in this portion of the section; they were not very abundant. Between this scrub and the creek, occur pied formerly by a scrub, not very dense, and in other parts by a marsh and an alpine, meadow, herbaceous plants are now a great feature, the whole tract being now occupied by a rich vegetation, conspicuous amongst which are huge patches of Celmisia armstrongii. Section D comprises that portion of the pass adjacent to Lake Misery and the old moraine, as well as the flanks of Mount Rolleston, joining at its southern limit Section O. The portion near the lake forms a most instructive portion of

the Midland Railway fire. Formerly the lake was enclosed by subalpine scrub, and the south face of the moraine was, also densely covered. The ground at the south end of the lake is rapidly becoming covered with fresh growth. In a patch of burnt Dracophyllum longifolium 20 cm. square were counted forty-five young plants of Celmisia armstrongii, and in 16 cm. by 8 cm. were thirty seedlings of Ligusticum haastii. The new scrub on this side will be, in large measure, Veronica and Cassinia; formerly Dracophyllum longifolium was the leading feature. On the western side of the lake is a piece of subalpine scrub left intact. This is divided into two halves, where the fire partially burnt out its middle portion, and so forms a splendid example, for one can tell for a fact of what the original scrub exactly consisted. The following is a list of the present living species, arranged in order of most frequent occurrence:— 1. Dracophyllum longifolium, Br.; Phyllocladus alpinus, Hook, f. 2. Coprosma parviflora, Hook. f.; Olearia nummularifolia, Hook. f.; Cassinia vauvilliersii, Hook. f. 3. Gaya lyallii. Jack, and Hook.; Dacrydium colensoi, Hook, f.; Panax colensoi, Hook, f. 4. Dracophyllum traversii, Hook, f.; very plentiful in the tallest, but not common in the lowest scrub. 5. Hymenanthera,* One of the species doubtless formerly included under H. crassifolia, Hook. f.; but, in absence of specimens, I am not able to refer it to its new species according to Kirk in “Transactions of the New Zealand Institute,” vol. xxviii., p. 510: “A Revision of the New Zealand Species of Hymenanthera, R. Br.,” by T. Kirk, F.L.S. sp.; scarce. 6. Veronica odora, Hook, f.; very scarce. 7. Clematis australis, T. Kirk; scarce. 8. Veronica subalpina, sp. nov.; very scarce. Especially noticeable is the small quantity of Veronica. Herbaceous Plants and Grasses. Phormium cookianum, Le Jolis; a few plants. Viola cunninghamii, Hook. f. Ourisia macrophylla, Hook, f. Danthonia raoulii, Steud. And perhaps others not noted. List of Plants Which Have Appeared in Burnt Portion of Section D Since the Fire. Coprosma serrulata, Hook. f.; in quantity. Senecio elæagnifolius, Hook. f.; in quantity. Cassinia vauvilliersii, Hook, f.; in quantity.

Dracophyllum longifolium, Hook, f.; in quantity. Veronica canterburiensis, J. B. Armst.; in large quantity. Veronica odora, Hook, f. Olearia nummularifolia, Hook. f. Coprosma cuneata, Hook, f.; from burnt stump. Aristotelia fruticosa, Hook, f. Gaya lyallii, Jack, and Hook. Phyllocladus alpinus. Hook. f.; one plant unburnt. Coprosma; ciliated variety. Podocarpus nivalis, Hook. f. Olearia illicifolia, Hook. f.; from old stump. Veronica salicifolia, Forst. Gaultheria rupestris, Br. Portions of this patch contain, very few seedlings, the ground being occupied by Coprosma serrulata, Lomaria procera, and Phormium. Regarding C. serrulata, see remarks in Section E. Although the plants are not coming up so closely as in many places, there is no reason to doubt but that in time they will form, a dense mass, unless the growth of Phormium should prove too much. Part of the new scrub at the base of the moraine is already one dense mass of Veronica subalpina and Cassinia, averaging 0.52 m. in height. At the margin of the lake, on the north side, is a flat patch of peaty ground covered with unburnt dwarf scrub, consisting of Veronica odora, 0.90 m. in height; Dacrydium laxifolium, Hook, f.; and Dacrydium bidwillii, Hook. f.; together with a few plants of Coprosma propinqua, A. Cunn. A continuation of this was burnt, and Veronica odora has since appeared in large quantities. Behind this is also another example of Veronica reproducing itself to the exclusion of all other shrubs. Proceeding up the face of the moraine towards its summit the following were noted as having appeared since the fire:— Veronica subalpina, sp. nov. (ined.); in large and close-growing patches, 0.60 m. in height. Olearia illicifolia, Hook. f.; at times. Senecio elæagnifolius, Hook. f.; plants here and there. Coprosma serrulata, Hook. f. Phyllocladus alpinus, Hook, f.; one or two seedlings noted, and these were among the very few seen during the taking of these notes. Olearia colensoi, Hook, f.; the only plant of this species noted in this locality. Panax colensoi, Hook, f.; Dracophyllum longifolium, Hook, f.; Phyllocladus alpinus, Hook, f.; Coprosma cuneata, Hook, f.”: a few plants spared by the fire.

In many places the reproduction is not very extensive. Here and there great numbers of seedlings of Dracophyllum longifolium were met with. In one spot, 0.18 m. by. 0.20 m., fourteen seedlings were counted. This species will probably occupy the upper portion of the moraine in time to come. Where the fire has passed over the adjacent slopes of Mount Rolleston (Canterbury), as far as the eye can reach is green with Veronica. On the wind-swept summit of the moraine the former subalpine scrub has been nearly eradicated, the only shrubby vegetation reappearing being Coprosma serrulata, together with herbaceous plants in the shape of large quantities of Celmisia armstrongii, Phormium, and Aciphylla colensoi. The moment, however, any shelter comes—for instance, a large rock or a slight depression in the ground—Dracophyllum longifolium in quantity, Senecio elæagnifolius, Veronica canterburiensis, and Olearia illicifolia make their appearance. With this section I include the slopes of Mount Rolleston, over which the fire swept in all its fierceness. These were formerly covered with a dense scrub similar to that described in this section, but wanting Dracophyllum traversii. I well remember before the fire forcing my way through it. It is now the home of vast numbers of herbaceous plants, from seedlings to those of full maturity. Especially conspicuous is Ligusticum haastii, Celmisia coriacea, Celmisia armstrongii, Ourisia macrocarpa, and Phormium, cookianum; indeed, here may be found representatives of nearly all the herbaceous flora of the district, except those plants which seem to require for their well-being a deep winter covering of snow, or those whose habitats are rocks at a high altitude. Amongst these plants shrubs are rapidly springing up, as pointed out in Section A, especially Coprosma serrulata, Coprosma ramulosa (in places), Cassinia vauvilliersii, and Dracophyllum longifolium. Where the scrub was formerly very thick Veronica subalpina, 0.45 m. to 0.60 m. in height, occurs in large, quantity, together with Coprosma parviflora (common), Gaya lyallii, Clematis australis (common), Aristotelia fruticosa (not so abundant), Coprosma acerosa (a little), and Gaultheria antipoda. Section E : This is of considerable size, and occupies all the right bank—south—of the River Otira from the West Coast Road for a distance of 1 kilom. up the river to an altitude of 1,000 m., from which point the ancient scrub is untouched, so far as its extreme altitudinal limit. This section was examined with considerable care, so far as time would permit, and is described at some length, in view of future observations, although a good deal is merely repetition of what has gone before.

Proceeding along the slope above the river in many places at first the original scrub was not very dense, and consisted in large part of Dracophyllum longifolium, and such shrubs as have been previously shown to be associated with it. These are not in general reproducing themselves. Coprosma serrulata has often taken complete possession; and of this species, strange to say, no young plants were observed. This creeping shrub must have great power of resisting fire, its strong underground stems apparently escaping injury. So soon as the fire is over it must recommence growth at a great rate, there being no longer the dense shade to keep it in check; and so, for a time, the sole survivor, it takes possession of the place, spreading entirely by means of its underground stems, their growth possibly stimulated by the potash as well as by the extra light, air, and direct rain. These remarks apply, though in a limited degree, to its more recently described relative C. ramulosa. These will probably in time to come remain almost; the sole occupants of the large areas they have seized upon. Proceeding up the river, here and there young plants of Dracophyllum longifolium were met with; also a sprinkling through the whole of Olearia illicifolia. Perhaps the great feature of this portion of the section is the vast number of seedling plants of Celmisia, especially C. armstrongii, which must have increased considerably in quantity since the fire. Less than two years after the fire I was over this same ground, and Celmisia was then germinating in great abundance. A few plants of Senecio elœagnifolius and Gaultheria antipoda were also noted. Proceeding up the river-bank a gully is reached where formerly existed a much taller and denser scrub. This—the species being arranged in their order of frequency—consisted of—Gaya lyallii, Phyllocladus alpinus, Senecio elœagnifolius, Dracophyllum longifolium, Olearia nummularifolia, all these in plenty; with, in smaller quantity, Coprosma cuneata, Veronica salicifolia, Podocarpus nivalis, and Senecio bidwillii. Here Veronica subalpina, 0.20 m. to O80m. in. height, is in great quantity, and, with Phormium and Coprosma serrulata, will replace the original totally different scrub. Other seedlings observed here, but in much smaller quantity, were: Veronica canterburiensis, Coprosma parviflora, Coprosma acerosa, Olearia illicifolia, Veronica salicifolia (here and there), Coprosma cuneata, Coprosma ramulosa, Coprosma retusa, Petrie, with here and there a few of such trees as usually are amongst those which escape burning. It is worthy of remark that no Veronica subalpina was observed in the still standing scrub. Passing over this gully, at first Coprosma serrulata takes precedence, and then a considerable quantity of Dracophyllum

longifolium and Veronica subalpina is met with, which must in time form a thick scrub. Wherever the ground gets moister Veronica odora makes its appearance in considerable quantity, 0.45m. to 0.52 m. in height, and already blooming. Further on still are remains of dense Dracophyllum, now replaced by the three Veronicas, 0.55 m. to 0.58 m. in height. One patch of Veronica subalpina, in area 5 m. by 1.60 m., measured in height 0.70 m. At about 1 kilom. from the road the burning ceased, a creek from Mount Rolleston proving a barrier to the fire. Across this creek the original subalpine scrub is encountered in its primeval state. Here it is so dense as to be practically impassable. Just before reaching this point the reproduction is not very fast, the young scrub consisting mainly of Veronica subalpina and Coprosma serrulata, together with Phormium in plenty, herbaceous plants, and grasses. Over the whole of this section Epilobiums have been frequent, three species being mainly observed. Epilobium, as a rule, is very abundant after fire, the burnt forest near Lake Wakatipu, on the Humboldt Mountains, abounding in varieties, many of which are possibly hybrids. List of Shrubby Plants in the Primeval Subalpine Scrub of Section E. (1.) Growing on Stony Ground and with Partial Shelter. Veronica subalpina, sp. nov. (ined.). Coprosrna cuneata, Hook, f. Senecio elœagnifolius, Hook, f. Coprosma, ciliated sp. Senecio bidwillii, Hook. f. Dracophyllum longifolium, Br.; not in quantity. Muhlenbeckia axillaris, Hook, f. Veronica canterburiensis, J. B. Armst. Olearia nummularifclia, Hook. f. Hymenanthera (vide note, ante). Species Where the Scrub Becomes Taller. Panax colensoi, Hook, f.; in quantity. Senecio elœagnifolius, Hook, f.; in quantity. Phyllocladus alpinus, Hook. f.; in quantity. Archeria traversii, Hook, f. And chose enumerated above in smaller quantity, and Veronica almost altogether absent. To sum up, there is on the ridges a tall Dracophyllum-Phyllocladus-Panax scrub, and in the hollows a dwarfer one of Veronica and various dwarf shrubs of which Olearia nummularifolia fills a large part. Very little Coprosma serrulata was seen, Podocarpus nivalis, and in places Coprosma ramulosa, being in its stead.

On the opposite—north—side of the River Otira the fire spread to the grass-meadow-line. Here is only one small patch of scrub to indicate the former vegetation, and this is of the usual mixed description common to high levels in West-land, and described in Section D. Veronica subalpina and Cassinia vauvilliersii are rapidly occupying the burnt ground, especially the former species. Of this portion I can give no detailed account, not having had opportunity to examine it carefully. Section F includes the slopes of Mount Rolleston (West-land), on the left bank of the River Otira, near the junction with Peg-leg Greek, and stretches south and west to Section E. It was here that the fire originated. Only the living scrub was examined carefully, the state of the River Otira not allowing the slopes on Mount Rolleston to be examined. There the chief plant seemed to be Dracophyllum longifolium. My notes say, “The scrub is being reproduced, but not, so far as I could see, to any great extent.” The constituents of the former scrub are identical with the list in the next section. Section G comprises Peg-leg Flat and all the adjacent slopes of Hill's Peak, as far as the slope opposite to Lake Misery. This was burnt by the first fire, the date of which I am unable to furnish; nor do I know whether the whole of this section was burnt at the same time. The evidence of the standing stumps and the state of the reproduced plants seems to point to more than one fire. It is also possible that in more than one place the Midland Railway fire may have reburnt some of this, or some may have been fired separately at or after that time. Be this as it may, an excellent example is shown of what takes place on the burnt ground after a considerable number of years have elapsed. Commencing at Peg-leg. Fiat is the living remains of a scrub burnt perhaps twenty years ago. This is more correctly designated low forest than subalpine scrub, and contains, as did all the original scrub of this section and Section F, in addition to most of the plants already treated of, various members of the forest at a lower level, and which here exist at their highest altitudinal limit. List of Shrubs Forming Low Forest, Near Peg-Leg Creek (in order of abundance). Dracophyllum traversii, Hook. f.; in large quantity. Gaya lyallii, Jack. and Hook.; in large quantity. Olearia illicifolia, Hook. f.; nearly as abundant. Panax colensoi, Hook. f.; plentiful. Dacrydium colensoi, Hook. f.; plentiful.

Phyllocladus alpinus, Hook. f.; plentiful. Senecio elœagnifolius Hook. f. Griselinia littoralis, Raoul. Coprosma parviflora, Hook. f. Panax anomalum, Hook. f. Podocarpus nivalis, Hook, f. Coprosma cuneata, Hook. f. Dracophyllum longifolium, Br. Myrsine pendula, Col. (?). Underneath this are large quantities of seedling plants, of which the chief are :— Gay a lyallii, Jack, and Hook. f.; in great variety. Senecio elœagnifolius, Hook. f. Veronica salicifolia, Hook. f. Aristotelia fruticosa, Hook, f. Coprosma parviflora, Hook. f. Griselinia littoralis, Raoul. Panax colensoi, Hook, f. The portion of the hillside adjoining this is now one mass of young growth. Seedlings of all kinds are present, also growth from burnt stumps. The following is a fairly complete list:— Senecio elœagnifolius, Hook. f.; very abundant. Coprosma parviflora, Hook. f.; abundant. Coprosma acerosa, A. Cunn.; abundant. Dracophyllum longifolium, Br.; abundant. Cassinia vauvilliersii, Hook. f.; abundant. Veronica canterburiensis, J. B. Armst.; small Seedlings. Coprosma serrulata, Hook. f. Gaultheria rupestris, Br. Veronica subalpina, sp. nov.; not abundant. Olearia nummularifolia, Hook. f.; not abundant. Gaultheria antipoda, Forst.; not abundant. Dacrydium laxifolium, Hook, f. Dacrydium colensoi, Hook. f.; only one plant noted. Podocarpus nivalis, Hook. f. Panax colensoi, Hook, f. The height of this new scrub is from 2 m. to 0.60m. In one place on the terrace facing south-west Dracophyllum longifolium is the leading variety, it having taken the place of a former, dense growth of Dacrydium colensoi. As Peg-leg Creek is approached the burning was only partial, and here, notwithstanding, a new growth similar to that just enumerated has made its appearance. At a higher altitude Dracophyllum longifolium, and Senecio elœagnifolius appear to be the leading shrubs. Descending the ‘terrace, and crossing Peg-leg Creek, on

the very steep slope which forms the commencement of its gorge is both living and burnt vegetation. The living consists of low forest 6 m. in height (Plate XXXVII.), of the following species:— Panax simplex, Forst. Griselinia littoralis, Raoul. Phyllocladus alpinus, Hook, f. Olearia illicifolia. Hook. f. Archeria traversii, Hook. f. Dacrydium colensoi, Hook. f. Gaya lyallii, Hook, and Jack. Panax colensoi, Hook. f. Olearia macrodonta, Hook. f. Dracophyllum traversii, Hook. f. The abundance or otherwise of this was not noted. Underneath the shade of these shrubs are seedlings of most of the above, together with Panax anomalum, Panax lineare, and Rubus schmidelioides, A. Cunn. Where the scrub was burnt on the steep face few seedling plants have appeared, the whole surface being invaded by a strong growth of Lomaria procera and Phormium cookharium. The following were noted:— (1.) Growing from Old Stumps. Coprosma serrulata, Hook. Olearia nitida, Hook, f. Gaultheria rupestris, Br. (2.) Seedlings. Coprosma cuneata, Hook. f. Olearia haastii, Hook. f. Olearia illicifolia, Hook. f. Veronica subalpina, sp. nov. Olearia avicenniœfolia, Hook. f. This face catches the full force of the north-west wind, and so in all probability the covering of fern and Phormium will hold its own. Between the unburned scrub and the West Coast Road, in an almost identical situation, reproduction is going on freely, Veronica salicifolia and Olearia illicifolia forming the main portion, and in such quantity as to be likely to exterminate the fern. Proceeding up the road until opposite to Lake Misery, and facing between west and south-west, is a portion of Hill's Peak, the scrub on which formerly consisted in the main-of :— Dracophyllum traversii, Hook. f. Dracophyllum longifolium, Br. Phyllocladus alpinus, Hook. f. Gaya lyallii, Jack and Hook. f.

Olearia nummmularifolia, Hook. f. Coprosma cuneata, Hook. f. This is now being reproduced in the form of— Veronica subalpina, sp. nov., about 50 per cent. Olearia illicifolia, Hook. f. Senecio elcœgnifolius, Hook. f. And a few plants here and there of— Panax colensoi, Hook. f. Cassinia vauvilliersii, Hook, f. Gaya lyallii, Jack. and Hook, f. On the slopes of the neighbouring mountains it is quite easy to tell of what the primeval scrub chiefly consists by the colour of the vegetation, a reddish-brown pointing out the presence of Dracophyllum, and a greyish-green that in which Phyllocladus is the leading tree. These various sections described give a fair idea of what the vegetation once was over the greater part of Arthur's Pass and its vicinity, and of what is now coming into its place, where destroyed by fire. Some of the sections I was, of course, able to examine with greater care than others, while a very wet season and other botanical work hindered my reporting on certain very interesting portions of the area. I ascended the mountains on both sides of the pass more than once to a considerable altitude, but, beyond noting that the fire reached to 300 m. above the pass in places, took no notes; such at the extreme limits of the fire should be of great interest. To some extent those taken at 1,000 m. up the Otira Valley supply this want. Looking over these notes, the following conclusions seem to be fairly justified :— 1. That subalpine scrub such as that in the Arthur's Pass district, after fire, is usually re-established, but in a different form to that which it had before the fire. 2. That the nature and amount of its re-establishment depends in large measure on the situation of the original scrub with regard to altitude, sunshine, and prevailing winds. 3. That Veronicas of various species often entirely take possession of places where formerly not a single Veronica existed. 4. That a Veronica scrub reproduces its like. 5. That Dracophyllum longifolium, Br.; Senecio elœgnifolius, Hook. f.; and Cassinia vauvilliersii, Hook, f., are also very readily reproduced after fire, and will form a large part of the future vegetation. 6. That fire in such a region as Arthur's Pass does little or no damage to herbaceous plants, grasses, and ferns; that, on the contrary, certain species, such as Celmisia armstrongii, Petrie, and perhaps Phormium, become more abundant.

7. That certain low-growing, creeping shrubs, notably Coprosma serrulata, Hook. f., and to a lesser extent C. ramulosa, Petrie, multiply by stolons to an astonishing extent, and become quite a new feature in the vegetation. 8. That certain shrubs specified in the foregoing notes, when burnt to the ground, readily spring up again from the old stock. 9. That the seedlings most frequent under the growing scrub are not the same species as those which appear in the open most abundantly after a fire. 10. That Dracophyllum traversii,* Of this no seedlings were noted, but on the more western mountains seedlings are not uncommon, in association with mature plants. Hook. f., is completely eradicated by fire; also that Dacrydium colensoi; Hook, f., and Phyllocladus alpinus, Hook, f., are almost eradicated. 11. That, owing to certain shrubs no longer playing a leading part in the vegetation, the colours of a mountain-side would be changed after reproduction of the vegetation. Why certain shrubs should not be reproduced, and others reproduce themselves so readily, or even usurp ground not originally their own, brings me from the region of fact to that of conjecture. Veronica, the most striking example of this, has always seemed to me a genus in which the species are not yet completely differentiated. † The species of Veronica as described by various botanists vary so much that no good key to the New Zealand species has as yet been published. In one gully on Mount Torlesse I have collected thirty quite distinct forms of Veronica odora, many of which are so distinct in appearance that a systematic botanist unacquainted with this variability would at once assign to them specific rank; and so with most of the other species as now known. The late J. Ball, F.R.S., held similar views with regard to Escallonia, Rubus, Hieracium, Solidago, and other genera. Writing of Escallonia, he says, “It is easy to find specimens not exactly agreeing with any of the described species, and to light upon intermediate forms that seem to connect what appeared to be quite distinct species. They afford an example of a fact which I believe must be distinctly recognised by writers on systematic botany—that in the. various regions of the earth there are some groups of vegetable forms in which the processes by which species are segregated are yet incomplete, and amid the throng of closely allied forms the suppression of those least adapted to the conditions of life has not advanced far enough to differentiate those which can be defined and marked by a specific name. To the believer in evolution it must be evident that at some period in the history of each generic group there must have occurred an interval during which species as we know them did not exist.”‡ Notes of a Naturalist in South America,” by John Ball, F.R.S. (p. 181).

If this be so, then a genus in such a phase seems to me to be at its greatest height of vigour, as evidenced in this case by Veronica taking entire possession of an alien's territory and easily repopulating its own. And the converse of this should hold true—that a very well defined species, confined to a certain restricted environment, will be very ancient and of feeble reproductive powers; such an one is Dracophyllum, traversii, found only at a certain altitude, with plenty of shelter provided by neighbouring shrubs, and only in positions of excessive moisture and great drainage. It would possess little vital energy, and so be ill able to reproduce itself under changed conditions. And so with all other species; so that by a series of investigations in various parts of our Island it might be shown which are modern and which more ancient forms, the comparative age of any species being so determined by observations made not merely after fire, but on prevalence of seedlings, vitality of seed, ease of culture, and the like. A fire thus may give us some idea of what may happen to the vegetation as the climate slowly changes, those forma which are comparatively young being the survivors, while the old will perish. Thus, from the tabulated results of such burning as this described here, we may get some idea of the future vegetation, shrubby Scrophularineœ, Compositœ, and Rubiaceœ in large measure taking the place of pines* Dr. Diels says, “Ferns, Coniferœ, Restiaceœ (probably), and certain Epacrideœ belong to the most primitive elements of the New Zealand flora (l.c., p. 292). (Translated.) and Epacrideœ.† Dracophyllum longifolium may seem an exception, since it has been reproduced in large quantities; but, as may be seen on reference to the various sections, its place is constantly being taken by other plants. Some bearing, too, my subject may have on the history of man in this Island. Fires having had their origin in early Maori times would still, perhaps, leave their mark in the shape of snbalpine scrub of new form, and the age of such would indicate the presence of man in that region at that date. Apart from the scientific significance of this subject, it is of interest from the layman's point of view, showing as it does that a large portion—the most beautiful portion, in fact—of our flora is not in great danger of eradication by fire; that if areas were set apart as national parks in the alpine and sub-alpine regions, and cattle and sheep kept religiously away, although an accidental fire might sweep over the whole locality, the loss to the colony and to science in the destruction of our unique subalpine flora need not be feared (see Plate XXXVI.).

Description of Plates XXXV.–XXXVII. Plate XXXV. Portion of unburnt subalpine scrub, with Aciphylla colensoi in centre, Veronica canterburiensis on left-hand side, and Phormium cookii on right hand. Part of Section A. Plate XXXVI. Herbaceous plants in Section C, reproduced after fire. Astelia nervosa, Ligusticum haastii, and Ourisia macrocarpa at bottom of plate; higher up Celmisia coriacea for most part. Plate XXXVII. Portion of Section G in foreground, and part of Section E with Canterbury, spur of Mount Rolleston in distance. Burnt and living sub-alpine scrub. Towards the centre of the plate Phormium and fern may be seen taking possession.