APPENDIX 



occur at intervals, and we observe here the Candle-nut Tree (Aleurites 

 moluccana), the Vunga (Metrosideros polymorpha), and the Thau-kuro 

 (Casuarina nodiflora). Such are some of the botanical features of these 

 districts ; but the reader will acquire a sufficiently correct general notion of 

 the floral physiognomy of these regions if he bears in mind their most 

 conspicuous characters, those of an undulating region more or less covered 

 with ferns, tail reeds, and grass, and dotted over, either separately or 

 in clumps, with Casuarinas (C. equisetifolia), Screw-pines (Pandanus 

 odoratissimus), Cycads (C. circinalis), and Acacias (A. Richii, &c). 



However, the peculiar vegetation of these plains often ascends the 

 lower slopes of the mountains, reaching to various elevations. In Y;mua 

 Levu it often ceases at 900 or 1,000 feet, but it may only reach to 400 or 

 500 feet, and, on the other hand, not uncommonly it ascends to as much 

 as 1,500 feet, the greatest elevation recorded by me being 1,600-1,700 feet 

 in the Sealevu district. It extends miles inland, and where conditions are- 

 suitable it may reach the heart of the island. 



Different explanations have been offered of the origin of the peculiar 

 vegetation of the leeward slopes of these islands. It is, however, a 

 phenomenon that is presented over much of the globe by islands lying in 

 the track of regular winds, the weather, or wet, side being densely wooded, 

 whilst the lee, or dry, side is covered with grass, ferns, and similar vegeta 

 tion. The predisposing cause must be climatic ; and although Mr. Horn 

 explanation attributing it to the effect of fires and to a faulty system 

 of native cultivation (pp. 80, 132) may be doubtless true in certain 

 localities, the influences at work here must be the same as are at work in 

 other islands and on continental coasts in other parts of the world. 



But for all that it is not easy to give a definite explanation even from a 

 meteorological standpoint. Those who are interested in this subject will 

 recall the desert districts of Australia and the dreary sandy wastes of 

 the coast of Northern Chile and Peru; and they will be cautious in 

 venturing on a definite explanation even with such relatively unimportant 

 examples of the same principle as are exhibited by the islands of Fiji 

 Dr. Seemann, writing of these " talasinga " plains (p. xih. remarks that 

 "their very aspect is proof that rain falls in only limited quantity,' 1 the 

 mountainous backbone of the islands intercepting, as he holds, much of the 

 rainfall. But the subsequent observations of Mr. Holmes, at I telanasau, 

 in the "talasinga " district on the north-west side of Yanua Levu, have 

 shown that there is by no means a small rainfall in this locality, the av< rage 

 rainfall, for instance, for the seven years ending December, 1 -^ 7 7 ■ being 

 113 inches, which must be quite two-thirds or three-fourths of the fall on 

 the weather side of the island (see p. 215) ; whilst the average number of 

 days on which rain fell was 156. The true cause would seem to lie in the 

 excessive dryness of the air on the lee side of the islands between the 

 rains, and the whole matter may, perhaps, be one rather for the hygrometer 

 than for the rain-gauge. I have no comparative data bearing on this 



