1893.] D. Prain — Flora of rTarcondam and Barren Island. 47 



appear to indicate that originally Narcondam may have been a volcano, 

 produced, like the volcano that appeared on the Island of Camiguin 

 in July 1S71,* by the extrusion of viscid lava without the accompani- 

 ment of crater-forming materials. In any case, the depth of the ravines 

 that plough the flanks of the hill on every side indicates very clearly 

 how remote has been the period of the island's activity.f 



The top of the island is frequently bathed in cloud ;$ during the 

 ten days spent in the island in 1891, this cloud-cap seemed to envelope, 

 for the greater part of the day, the last 400 feet of the peak. The appear- 

 ance, however, was slightly deceptive ; for it was noticed that the cloud 

 was only condensed on the western aspect of the hill, and that towards 

 evening the peak always became clear. The nature of the vegetation 

 on the peak, — the trees bearded with moss, and their bark covered with 

 Trichomanes — indicates clearly that this is a usual state of affairs. 



Save on the sea-cliffs, which are bare, and on the eastern side of 

 the peak near the top, where the jungle is thin and scrubby, the whole 

 island is clothed with dense forest : this consists mainly of lofty trees, 

 with but few climbers, in the beds of the various watercourses. On the 

 intervening ridges the vegetation consists of a tangled mass of shrubby 

 growth overloaded with creepers. Landing at Anchorage Bay one finds 

 on the shingle some plants of Ipomoea biloba ; immediately behind the 

 shingle, and under the shade of about a dozen coco-nut trees, is an at- 

 tempt at a sea-fence, composed of Sccavola Koenigii, Hibiscus tiliaceus, 

 Morinda bracteata, Guettarda speciosa, Pandanus odoratissimus ; some 

 Ipomaia grandiflora, Convolvulus parviflorus, and Wedelia scandens climb 

 over these. Behind these bushes some trees of Barringtonia speciosa, Termi- 

 nalia Catappa, Erythrina indica, Sterculia rubiginosa, Thespesia popidnea, 

 Dracaena angustifolia, Ardisia humilis, and Ixora brunnescens represent the 

 beach-forest. There is, however, but scanty room for species of either 

 class, and a few plants of Eranthemum succifolium underneath the trees 

 complete the representation of this sort of vegetation in this situation. 

 To the south of this point are some low cliffs, covered at the top with a 

 tangled mass of Hoya orbiculata, while at their base plants of Pluchea 

 indica, Blumea glomerata, Vernonia divergens, Desmodmm polycarpon, 

 Cyperus pennatus, and Tliysanolccna acarifera occur ; the last-named, — it 

 is, by the way, the only grass that is found on the island — is the most 

 plentiful and seems to be, besides Fimbristylisferruginea and Boerliaavia 



strongly supports the conclusion (to which Mallet also inclines) that there never 

 was a crater in Narcondam, and that the island is of the endogenous volcanic type. 



* Moseley : " Notes by a Naturalist on tho 'Challenger,'" p. 409. 



t Mallet : Memoirs of the Geol. Survey of India, xxi, 281. 



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

 J. ii, 7 



