518 A NATURALIST ON THE "CHALLENGER. 



>> 



soon built a small waterproof hut, with sticks and the huge 

 wild Banana leaves. Then they put up another small roof of 

 leaves, and finding dry dead Banana leaves under the shelter of 

 the freshly fallen ones, soon lighted a fire under the roof, and 

 we dried our clothes in the smoke before nightfall, in the midst 

 of the heavy rain. The Banana leaves afforded further water- 

 proof covers for our clothes and for my botanical drying paper. 



We had brought no blankets with us, because I wished to 

 make the utmost attempt to scale the mountains as far as 

 possible, and had therefore reduced the baggage to a minimum. 

 I had not expected that we should suffer from cold as we did. 

 The thermometer showed, at about half an hour before sunset, 

 75° F., about an hour later, 68°'5, at midnight 63°-0, at daybreak 

 60°'5, and in about half an hour after daybreak it rose to 61-5°. 

 The main stream of the valley running past the huts, had a 

 temperature at daybreak of 65 o, 0, having retained throughout 

 the night the heat of the former day, which the air had so 

 rapidly lost. The effect of the stream on the climate here, is 

 thus just the opposite of that of the streams of such an island 

 as Tristan da Cunha.* 



From this camp, the way led over several steep minor ridges 

 in the head of the valley, and then up to an elevation of 3,000 

 feet, which was reached on one of the extremely narrow ridges, 

 characteristic of Tahiti, situate just to the west of the u Diadem." 

 From the ridge, a descent was made into the Punaru Valley by 

 the aid of ropes fastened to the trees. The precipitous side of 

 the valley which we thus descended, was covered at this eleva- 

 tion, from about 3,000 to 2,000 feet altitude, with a dense 

 vegetation, composed almost entirely of ferns. A Tree Fern 

 (Alsophila Tahitiensis) formed a sort of forest to the exclusion 

 almost of other trees, and with this were associated huge clumps 

 of the giant fern, Angiopteris evecta, and masses of the Birds-nest 

 Fern (Asplenium nidus). With these grew a trailing Screw- 

 pine and a Draccena, but the three ferns together formed a 

 greater proportion of the entire vegetation than I have observed 

 to be the case elsewhere.t 



* See page 111. 



t This statement concerning the preponderance of ferns in the vegeta- 



