44 D. Prain — Flora of Narcondam and Barren Island. [No. 2, 



The chief interest of this configuration resides in the misapprehen- 

 sions as to the structure of the island to which it Las given rise. 

 McCelland mistook either the ravines or the ridges between them for 

 streams of lava* ; Kurz has described and figured the island as a central 

 volcanic cone, surrounded by an outer ring, not much over half the ele- 

 vation of the central mass, and very largely broken down.f Seen from 

 Kurz's point of view (N. W., \ E"., at a distance of 20 miles) an oblique 

 view of the mouth of the yawning south-western gorge is obtained, while 

 the main mass hides the connection of its southern wall with the cen- 

 tral peak. At the same time the peaks already mentioned as connect- 

 ing the main hill with the somewhat outlying north-eastern spit, serve 

 to conceal their own connections and complete the illusion. At this 

 distance too the three hummocks at the top of the peak look very much 

 like as many points on the edge of a crater. In a nearer view from 

 the same direction the appearance of a central cone is still well- 

 preserved, though the regularity of what seems at a distance the 

 remains of an outer ring quite disappears. J Even close in-shore it 

 is impossible to say whether the three points on the peak are, or are 

 not, indicative of the remains of a crater, the forest that clothes them 

 disguising their true relationship. The appearance from another 

 point of view (W. \ S., at a distance of 40 miles) agrees well with 

 the description by Horsburgh of " a cone or pyramid with its sum- 

 mit broken off."§ 



* McClelland : Jour. As. Soc, Beng., vii, 77. It would depend a good deal on the 

 distance from which the island was seen, whether the ravines or the ridges between 

 them be what were taken for ' lava-carrents.' Seen from a distance of 6 miles or 

 more, through a glass, the darker shadows caused by the gorges might well enough, 

 as Ball (Records, Geol. Survey of India, vi, 89), and Mallet (Memoirs, G-eol. Survey 

 of India, xx, 281) suppose, be what led McClelland astray ; as however the drawing 

 on which McClelland based his opinion was taken from about a mile and a half, or 

 two miles from the eastern shore — the drawing was made by Griffith — there is 

 no doubt that what he took for streams of lava were the ridges between the 

 ravines : on this side of the island these are, towards the top, bare and rugged, and 

 are not unlike streams of lava. After all, however, McClelland had nothing to 

 support his idea that the island was volcanic but its conical shape and its isolation. 



t Kurz : Report on the Vegetation of the Andaman Islands, p. 4. Kurz appears 

 to have had nothing more to go upon in supposing the island to be volcanic than 

 had McClelland ; the accident of configuration led him to go further than McCelland, 

 and assume, not only that the island is volcanic, but that it is an island of the same 

 type as Barren Island, in which there is an inner and an outer cone. And with the 

 accounts and the appearance of Barren Island in his recollection — Kurz disposes 

 cursorily of Barren Island in the sentence immediately preceding the one referred to 

 — the idea is by no means unnatural. 



X Ball : Records of the Geol. Survey of India, vi, 89. 



§ Horsburgh: Indian Directory (ed. v), ii, 50. 



