Ecological Botnnij.] SUBANTARCTIC ISLANDS OF NEW ZEALAND. 207 



of the leaf, which otherwise is made up of a lioinogeneous tissue of ch:)se rounded 

 cells. I'lmrophijlluut speciosum has no cuticle. Numerous jointed hairs are de- 

 veloped from epidermal cells of both surfaces. Beneath the upper epidermis is a 

 2-layered water-tissue of large colourless cells, and beneath these is the chlorenchyma. 

 made of square cells (in section) of uniform size, 2 layers having the cells close and 

 acting as palisade, and the remainder in superimposed vertical rows separating large 

 air-cavities. The lower epidermal cells are very small, and give rise to numerous 

 hairs. The rilis are made of rounded collenchyma. Helichrysum beUidioides (creeping 

 sulfruticose plant of wettish ground) has a thin-walled epidermis of large colourless 

 cells, 2-layered palisade of rather short cells, the second row the shorter, large-celled 

 water-tissue in centre of leaf, and small-celled open pneumatic tissue with abundance 

 of chlorophvll. The lower epidermis is small-celled, with dense mass of hairs pro- 

 tecting the numerous stomata.* 



5. The Plant Formations. 



(A.) GENERAL. 

 With regard to the plant formations, those of each group of islands are con- 

 sidered separately, except those of Campbell Island, which are dealt with by Mr. 

 Laing in a separate memoir in this volume. Ecologically certain of the formations 

 of the different groups are identical, while floristically the differences are but slight. 

 The Danthonia meadowsf of the Auckland and Campbell Islands are the same, and 

 so are the Olearin LijnUii formations of the Snares and Ewing Island ; so, too, thougli 

 differing floristically, the Pleurophi/Unm Hookeri formation of the subalpine zone 

 of Auckland, Adams, and Campbell Islands is ecologically identical for all the islands. 

 As with most plant formations elsewhere, these under consideration merge gradually 

 into one another, and the description refers to what I take to be characteristic 

 portions. It will be seen that the " Pleurophyllum meadow " is a much more 

 uncommon formation than I had previously conjectured (Cockayne, " Botanical 

 Excursion," p. 257), while the Pleurophyllum. Hookeri formation of the Aucklands 

 has not been described previously. 



(B.) THE 8NARES. 

 (a.) General. 

 Seen from the sea are precipitous cliffs, white in ^jlaces with guano, and marked 

 here and there with yellow and green patches of vegetation. Above is a low, close 

 covering of greyish-looking shrubs, which give the special character to the scene ; 

 while in the narrow space on the upper margin of the cliffs is a belt of grass with 

 occasional bright-green patches of Veronica elUpticn. A closer view shows the general 

 grey colour to be modified here and there at its outskirts, especially near the sea, 

 by small green jaatches denoting the presence of Senecio Stewartiae. A more detailed 

 examination proves that the vegetation is not so uniform as appears at first sight, 

 and that, though the indigenous species number but twenty-three, climate, soil, 

 and birds have sorted them out into well-defined plant formations. 



* For special details re anatomy and numerous figures see Herriott, Trans. N.Z. Inst., vol. xxxviii, 

 p. 377 ; 1905. 



t The meadows miulit iivubaliK- Ijc ternKnl " moors," according to W'anninu's classific-ation. 



