fast killing out the species of Panax, which are completely barked up to 4 ft. or 5 ft. from the ground. Horses will also bark the shrubs of this genus.

The Waipakihi River (the main source of Lake Taupo and the Waikato River) rises in the Kaimanawa Mountains, and flows through them for some eighteen miles before emerging into the plain. The upper reaches drain some thousands of acres of pumice-flats, the Kaimanawa Range, although of old sedimentary formation, having been plentifully peppered with pumice from the contiguous voleanic area. The pumice-flats contain material in which rabbits may easily burrow, and they have accordingly taken possession in their thousands. The flats are at an elevation of from 3,000 ft. to 3,500 ft., and are covered with a shrubby growth of Veronica buxifolia, V. laevis, V. tetragona, and V. salicifolia (family Scrophulariaceae), or with tussocks of Poa caespitosa and Festuca rubra and Danthonia Raoulii (family Gramineae), with smaller plants between, such as Raoulia australis, Acaena sp., &c., and Aciphylla Colensoi. It is difficult, owing to the ravages of the rabbits, to say what the flora of the grass-flats originally was. Probably a number of finer grasses have been eaten out, and evidence is not wanting that food is scarce. Little piles of the leaves of Veronica salicifolia were seen near the bushes, which may be attributed to rabbits nibbling the stalks oft and leaving the blade. Many plants of Aciphylla Colensoi (family Umbelliferae) had been eaten down to the heart. The harsh Festuca rubra tussocks seemed to have suffered less than other grasses, but even these had been occasionally attacked. At Waipahi (Kaimanawas), just above the bush-line at 4,400 ft., on the hillside, Danthonia Raoulii was the sole survivor, and that had been badly eaten, possibly by wild horses as well as by rabbits and pigs.

These two valleys, though widely separated, are excellent examples of natural sanctuaries, including extensive river-flats walled in by steep, heavily bushed mountains, the only natural outlet being a long, winding gorge where the river issues into the plain. The floras of these valleys are rapidly changing in character owing to the attacks of imported animals, which, because of difficulty or illegality of access to the valleys, are not kept in check by man.

One effect of imported animals may be to restrict the more edible plants to situations beyond their reach. A species, therefore, which is able to adapt itself to any station may by compulsion be restricted to one. For instance, Lepidium oleraceum (“Cook's scurvy-grass”—family Cruciferae) has been eaten out along the Wellington Coast, and is now generally only to be found growing on inaccessible rock-faces. Similarly, Senecio Greyii (family Compositae), although able to grow on any soil, as testified by its presence in most collections of native shrubs, at Mukumuku, Palliser Bay, is restricted, possibly chiefly owing to goats, to stations which would lead the ecologist to class it as a chasmophyte. One may see abundance of this beautiful free-flowering plant growing on the cliffs, but it is with great difficulty that specimens may be secured. A common plant of littoral rock-faces is the grass Agropyrum scabrum, which is greatly relished by sheep. It is being eaten out possibly on the central volcanic plateau of the North Island and elsewhere. At Alexandra, Central Otago, this grass assumes the habit of a tussock-grass, and is then better able to resist close cropping.

It is, indeed, on the littoral that the evidences of the destructive influence of animals on vegetation are most readily found. The long winding coastline of South Wellington Province affords a commonage where both domestic

and wild animals may resort for salt-licks and saline plants, such as Salicornia australis and other plants of the salt-bush family (Chenopodiaceae). When food is scarce the giant perennial-stemmed grass Spinifex hirsutus is voraciously eaten, as has been observed at Pencarrow Heads and Titahi Bay. At the latter habitat marram-grass (Ammophila arundinacea) was untouched, but horses had eaten the Spinifex off short. The preference which stock exhibit for Spinifex over marram should be taken into account in considering the rival merits of the two grasses as sand-binders. A fern, Gymnogramme, once abundant on Wellington coasts, is believed to have been exterminated by sheep and rabbits. Being an annual it would be eaten before the spores were shed.

At Rocky Bay (Titahi Bay) is a recently raised beach. High above high-water mark is a boulder beach, then a sandy strip containing freshwater pools, the sand being covered with a closely cropped sward of (1) Crantzia lineata (family Umbelliferae), (2) Samolus repens* (family Primulaceae), (3) Ranunculus acaulis, named in the order of their relative abundance. There are also present Atropis stricta (family Gramineae), Cotula coronopifolia (family Compositae), and Selliera radicans (family Goodenovieae). Sheep greedily browse on this sward and drink from the pools, in which frogs are living. The term “salt meadow,” which is applied by ecologists to this formation, must therefore not be interpreted in a superlative sense. Separating the salt meadow from the sea is a raised rocky terrace. Sheep have been observed browsing on a sward of similar composition near Island Bay. On the dry hillside above Rocky Bay Eryngium vesiculosum (family Umbelliferae) has been closely eaten down by sheep, and it is feared that Lepidium tenuicaule var. minor (family Cruciferae), common here in 1907, has been entirely eaten out. On the rocky scarps near here Aciphylla squarrosa was observed in March, 1908, to have been badly eaten back. Further round the Titahi Bay peninsula, at the point facing Plimmerton, are stretches of raised sandy beach containing the remains of sea-animals and consequently much carbonate of lime. This has resulted in a shallow black soil supporting a sweet herbage, largely the naturalized alfilaria (Erodium circutarium, family Geraniaceae). The rabbits are spoiling much of this by covering it with earth from their burrows.

Changes in the indigenous flora by means of the spread of exotic species, the seed of which is distributed by imported animals, are being brought about in various localities. Examples which might no doubt be added to are the African box-thorn (Lycium horridum), which is spreading in the Taranaki bush; the blackberry (Rubus fruticosus), in many parts of the North Island and in the Nelson Province; the elderberry (Sambucus niger), near Dunedin; the gooseberry (Ribes grossularia), in many parts of the South Island; the Cape gooseberry (Physalis peruviana), in the Wanganui, Thames, and Tauranga districts; the ink-plant (Phytolacca octandra); and even the strawberry (Fragaria vesca), in some parts of the Auckland Province. All these fruits are spread by birds, especially the blackbird, and thereby the native vegetation certainly is being displaced.^†

Rats (both the grey and the black), by eating the seeds, undoubtedly influence the spread of many species. The grey rat has a fondness for the seeds of the New Zealand passion-flower (Passiflora tetrandra), the fruit of the kiekie (Freycinetia Banksii); and nikau-palm (Rhopalostylis sapida). Possibly the introduced birds may assist in the spread of the indigenous plants having edible fruits, such as the wineberry (Aristotelia racemosa), Fuchsia excorticata, poroporo (Solanum aviculare), and bramble (Rubus australis).

No one who has seen sheep covered with Acaena “burrs” (piripiri) can hesitate to admit the large part which that animal plays in the spread of this native weed.

Pigs are most partial to the seed of the hinau (Elaeocarpus dentatus, family Tiliaceae), on which they fatten, and the roots of bracken (Pteris aquilina). The native arrowroot fern (Marattia fraxinea) is fast being killed out by pigs on account of its large starchy rhizome.

The spread of clovers and other leguminous seed by animals must be a considerable factor in altering vegetation. At Palliser Bay gravel-fans, covering in some cases many acres, are formed by heavy rainfalls. The first plant to establish itself on the finer detritus is Raoulia australis, forming large depressed patches. Ultimately a certain amount of organic matter is formed by these patches of vegetation, and in October, 1907, clovers and other introduced leguminous plants were noted to be growing out of these patches. It is possible that they may in time displace the Raoulia, as in older but similarly formed land in the vicinity a close sward of Leguminosae monopolizes the soil.

The partiality which stock exhibit for certain shrubs such as the mahoe or hinahina, the so-called “cow-tree” of the settler (Melicytus ramiflorus, family Violaceae), the karamu (Coprosma grandifolia and C. tenuifolia, family Rubiaceae), and the broadleaf (Griselinia littoralis, family Cornaceae) have led to the practice among stockmen of cutting the shrubs down for fodder in times of scarcity. Another shrub evidently much relished is the mangrove (Avicennia officinalis, family Verbenaceae). Travellers along the Thames railway-line may see the lower branches of the fine mangrove shrubs of the estuaries trimmed off by cattle in the same way as they do the weeping willows in the meadows, no branches appearing below a certain level— the limit of the cattle's reach. In the Rotorua district I am informed that tawa (Beilschmiedia tawa, family Lauraceae) leaves are readily eaten by stock. Cattle greedily eat karaka-leaves (Corynocarpus Iaevigata, family Anacardiaceae). The fruit, poisonous to some animals, causes in pigs only a partial paralysis of the hind legs.

Central Otago, which contains such excellent examples of the effects of overstocking and of the rabbits on the indigenous vegetation, has furnished me with a few notes. On the Rock and Pillar Range (Middlemarch side) in December, 1908, Hymenanthera crassifolia^* (family Violaceae) was found to be eaten down by rabbits. (This effect has also been largely observed on the raised beaches at Turakirae Head, Palliser Bay.) Celmisia Lyallii (family Compositae) was almost eaten out except on a few inaccessible rocks at 3,700 ft. At Gimmerburn, on a dry hillside above the Government nursery, the ground was bare save for a few scattered plants of Agropyrum pectinatum (naturalized) and Koeleria Kurizii. These two grasses had been nibbled down very short, but were surviving, and the former was producing seed in quantity.

The soils of the Southern Islands are the very antithesis of those of Central Otago, but even here the flora is being slowly changed by imported animals. On Auckland Island, in November, 1907, at Flat-topped Mountain, Carnley Harbour, and above the scrub-line, pigs had eaten freely of Pleurophyllum Hookeri (family Compositae), having grubbed up the plant to get at the rootstock. At Port Ross, Auckland Island, in January, 1909. at 1,100 ft. numerous pig-tracks were observed, and Pleurophyllum speciosum appeared to have been eaten out on all stations but inaccessible rock-faces. At Enderby Island the cattle had considerably cut up the bush but their greatest effect was noticeable on the tussocks of Poa littorosa, a grass which is evidently being exterminated at that habitat. At Campbell Island, which is inhabited and farmed as a sheep-run, the Stilbocarpa polaris (family Araliaceae) is being eaten out by sheep. On the other hand, there are no pigs, as at Auckland Island, to attack the Bulbinella Rossii^* (family Liliaceae), which is spreading at an alarming rate. On the “burn,” under the “Judge's Chair,” at 725 ft. the Dracophyllum scoparium and D. longifolium are being replaced by Poa littorosa and Bulbinella Rossii. In one paddock near the wool-shed at Perseverance Harbour the flora consisted solely of a thick mass of Bulbinella Rossii—a magnificent sight.

The important family Leguminosae is sparsely represented in New Zealand. Horses are fond of chewing the taller species of Carmichaelia. It would be interesting to learn whether the dwarf species are diminishing owing to the attacks of rabbits in the more arid parts.

Sheep, and to a less degree cattle, frequently eat the smaller species of tutu (Coriaria, family Coriariacae), though it is not likely that they appreciably affect its abundance. Fern (Pteris aquilina, family Filices) is kept in check by close feeding by cattle. Both of these are instances of plants which under certain conditions may have a poisonous effect on stock, serving as part of a regular ration. The hinahina (Melicytus ramiflorus) has been suspected of injuriously affecting stock at Catlin's (Otago), but definite information is wanting.

Cattle and sheep, though having their preferences, will eat most ferns, any shrubs, and seedlings of forest-trees when food is scarce. In small isolated clumps of bush the undergrowth is sometimes completely destroyed.

Omitting the grasses, the native plants which find most favour with herbiverous animals would appear to be those belonging to the families Cruciferae, Umbelliferae, Araliaceae, Violaceae, Malvaceae, Tiliaceae, Rubiaceae, Primulaceae, Leguminosae, and Chenopodiaceae; Juncaceae and Cyperaceae contain genera (Juncus, Luzula, and Mariscus) species of which are often devoured by stock; Compositae and Liliaceae contain some species which are often browsed.

There are doubtless other non-poisonous native plants which supply either normally or in times of scarcity food for animals, and the author would be extremely obliged if other observers would publish or communicate to him any facts which will throw further light on the subject.

Very little has been recorded in the past on the subject of this paper. An interesting article on “The Displacement of Species in New Zealand,” by the late T. Kirk (Trans. N.Z. Inst., vol. 28, p. 17) gives a general account of the naturalization of many forms of life and the probable effect on the indigenous forms. Of special interest under the title of this paper are the