AFRO-INDIAN REGIONS. 



447 



be pumice; wafted into the air somewhat in the manner of froth; but 

 I had no opportunity of verifying tliis conjecture. 



I have ahxady spoken of single fibres of the cainllary obsidian 

 conveyed over great distances by the wind. This substance occurs 

 in tufts and quantities only beneath the cloud of smoke-like vapor ; 

 on an area extending beyond the limits of the pumice-tract ; collect- 

 ing and filling the inequalities and shallow depressions of the barren 

 lava-plain. On taking up a tuft, the threads were found to vary in 

 fineness; the coarser ones remaining straight, but the finer ones 

 crimped like fine wool, and often terminating in a diminutive drop or 

 globule of glass. Their origin has been traced to the vitreous lava- 

 crust, drawn out in fibres in the liftings and overturnings of the flow- 

 ing lava, when cooling in furrows that resemble those made by a 

 plough. Small quantities of tlie fibres are procured underneath such 

 projecting layers of lava ; a situation from which, they coidd hardly 

 be wafted out of the Crater : I was therefore led to suspect, that the 

 molten lake and its spouts might prove an additional source ; but no 

 direct confirmation could be detected, even on my nearest approach. 

 Mr. Drayton met with capillary obsidian "in the furrows of the lava 

 of the recent Eruption, as far as tlie place where it entered the sea :" 

 on the other hand, I looked in vain for this substance in the lava of 

 the terminal crater of Mauna Roa. 



It was not quite sunset when we passed the portion of the brink 

 nearest to the molten lake beneath ; whose surface was already dis- 

 tinctly illuminated. In company with others, I revisited this spot 

 later in the evening ; and looked down upon a spectacle, unique, pro- 

 bably, upon this planet. From the great amount of light emitted 

 the first feeling of surprise was, that only a small portion of the lake 

 should be luminous. We were looking down upon a circumscribed 

 space, which I could only compare to an immense slab of black mar- 

 ble veined with a network of lightning : but there were no sharp 

 angles in the lines of light, which remained permanent, and did not 

 pain the eye ; in intensity therefore, hardly exceeding the light of a 

 glass-furnace. The network, though always present, was always 

 moving; and in one di ei tion, regularly nnd slowly to the Southwest; 

 the meshes enlarging m the way, for they were not solid floes, and 

 coming as sharply against the bank, as though passing underneath. 

 The spouts at the farther end of the lake were altogether luminous, 

 constituting the most brilliant feature in the display ; the resulting 



