104 H. J. JOHNSTON-LAVIS ON THE GEOLOGY 
into two great groups dependent on their characters as rock-form- 
ing materials, and on their resistance to denuding influences. The 
first of these, both in point of age, density, and durability, are the 
lavas and scoriz of the ancient active volcano. As a rule, the 
most spongy of lapilli and scorie have a specific gravity above 2, and 
are often so closely blended together. either by pressure, decomposi- 
tion, or the infiltration of calcareous matter, as to be rendered 
capable of resisting the disintegrating subaerial influences. Even 
were this not the case, the great openness of structure en masse 
would allow the water to sink by the interspaces between each frag- 
ment and its fellow, and to gradually issue at a lower level in the 
form of springs, instead of rushing impetuously over the surface, and 
carrying materials away with it in its course to lower levels. 
The lavas, again, are threaded through and through with con- 
traction-jointage, but are so irregular, that one piece is as it were 
dovetailed into its fellow, so that we are often struck with the 
amount of molecular erosion that has gone on at the surface with- 
out being able to break up a comparatively thin flow. 
When, however, we come to consider the other group of materials 
which overlie the first, that is to say, the pumiceous tufas and ash- 
beds, the products ejected during the paroxysmal phases, characters 
are detected quite opposite to those of the first series of deposits. 
Loosely thrown together, with little or no binding material, fine 
dust often prevailing over larger fragments, we have large quantities 
of pumiceous rocks that, taken in mass, are much lghter than 
water. Again, owing to the much closer texture of the ash-beds, 
‘they offer greater resistance to water-sinkage, although readily be- 
coming saturated and again giving up their water and drying. It 
is Just these characters that account for the incredibly rapid denu- 
dation going on, so that we can actually measure differences from 
month to month. 
The average rain-iall at the Vesuvian Observatory is 780 milli- 
metres. This, although not a very extraordinary quantity, produces 
very evident effects by the manner in which it falls. For instance, 
not a drop may make its appearance for two or three months, fol- 
lowed sooner or later by a downpour, lasting from twelve to thirty- 
six or more hours, and having quite a tropical character. This large 
mass of water falling on the mountain slopes must necessarily be 
rapidly got rid of. 
The first thing, of course, will be to saturate the tufas; this 
accomplished, they may be considered as practically impermeable, 
at any rate to such a large bulk of water as is suddenly poured 
upon them. This is aided also by the vegetation, which is much 
richer on these pumiceous deposits. The water must therefore run 
over the surface, which, being highly inclined, gives to that liquid 
great rapidity of motion. The pumice, from its lightness, is rapidly 
swept along, leaving the denser lava-fragments without support, 
and these therefore are obliged to follow. When the current is deep 
enough, especially in the bottom of the gorges, a curious spectacle 
often presents itself, namely, a moving river of stones ; that is to 
