ON THE VOLCANIC PHENOMENA OF VESUVIUS. 317 



In a short time, with suitable apparatus, I collected over two kilogrammes 

 of this material absolutely free from mechanical impurities. 



Along many of the cracks of the lava beautiful glassy crusts of halite, 

 more or less impure, were formed, and often showed a dull red heat in 

 daylight. These crusts on being removed become rapidly opaque and milky 

 in hue, and audibly cracked into starch-like columns, due to the rapid con- 

 traction on cooling — producing, in fact, a miniature basaltic structure. 



About February 5, 1894, the lava was issuing in very small quantity, 

 and by the 7th showed no trace of movement. Yet even in May cracks 

 in the lava near its point of exit were incandescent some distance in, and 

 the saline incrustations mentioned above were in full perfection. 



Coincident with the ai-rest of the lateral outflow, the lava rose in the 

 chimney and the red reflection from the top of Vesuvius that had been 

 absent for so long, with rare exceptions, was again almost daily visible. 

 The level of the lava in the main chimney soon rose to the bottom of the 

 new crater that had been forming, and increasing in size during the time 

 the lateral issue of la\a had been going on, and commenced the tilling up 

 of that cavity by the formation of a cone of eruption, so that almost coin- 

 cident with the arrest of the leakage of lava laterally the central activity 

 changed from the crater- and dust-forming stage to the lava cake- and 

 cone-forming stage. 



I made a careful examination of the summit of Vesuvius about the 

 middle of May. The crater in an east and west direction was about 150m. 

 in diameter, and its depth, then decreasing, was about the same. The 

 walls were remarkably steep, in some places even vertical or overhanging. 

 The bottom could be seen with difficulty owing to the crumbling nature 

 of the edges. The walls are nearly all covered by sublimates or dust 

 that has adhered and crusted them over, so that several dykes both .solid 

 and hollow can no longer be distinguished. This is especially the case 

 with the one formed during the 1891 outburst. The details of the great 

 rift of the 1880-81 and subsequent eruptions on the east side of the great 

 cone were still easily discernible. On the south side, and a little to the 

 east, a wall of rock stands out from the side of the crater and is directed 

 nearly towards the centre. It is capped by a pinnacle of rock, and is really 

 the old dyke of the 1 885 eruption. 



Just to the east of that wall, and partly owing to its existence, the slope 

 of the inside of the crater is less in that direction. Here the guides had 

 made a little path for a few metres down. On examining carefully the 

 condition of things from its lower termination, which so far aided little 

 the view of what was going on at the crater-bottom, I found that by ex- 

 tending it down a slope, and then cutting a ledge farther round to the 

 east at a suitable point, a bracket-like platform some metres square could 

 be reached, which is about half-way down the crater. Later the path was 

 further widened by me and made more commodious, and now gives easy 

 access to the platform from which one can look right into the vent of the vol- 

 cano and watch with ease the boiling up of the lava and the ejection of the 

 great blobs and cakes that are rapidly filling up the crater. Unfortunately, 

 owing to the well-like shape of tlie crater, the shadows due to the vapour 

 column spreading out overhead, and the dark colour of the rocks, instan- 

 taneous photography could not be utilised to record this interesting and 

 ever-changing scene. 



As is usual at some period after an eruption, feathery gypsum is a 

 common product in the cavities of the old scoriae, and is associated at the 



