Jounston-Lavis—The Eruption of Vesuvius in April, 1906. 155 
termination of the volcanic duct, the cone in fact, is so incoherent that rarely is it 
able to withstand the out-thrust of the fluid column that it encloses, and a lateral 
outlet is formed. The overflow will be proportional, therefore, to difference of 
level between the top of the lava-column and the lateral outlet; and, in fact, where 
other disturbing factors do not interfere, general experience shows this to be the 
case. When such a lateral outlet exists, the primary overflow (by which I mean 
that due to the steady accession from the prime source) will drain off that excess. 
When the lateral opening is first established, in addition to this excess there will 
also be the fluid rock occupying the chimney above the level of the side tap, so that 
when this is established there is a big outpour, followed by a constant dribbling. 
To be more concise, the output from a lateral rent depends on :— 
(a) Amount of lava above the lateral outlet. 
(6) The secular output of lava. 
(c) The rise of magma due to its expansion from increased yesiculation 
after the relief of pressure from the fluid column above it has 
drained away. 
_ This was, in fact, the actual condition that so frequently prevailed during the 
period from 1872 to 1906, and, of course, has been the dominant characteristic of 
Vesuvius for some centuries. It is to this mechanism that the lava-cones, bosses, 
and buttresses of 1883, 1885, 1891, 1895, and 1903 owe their existence. This 
outflow is not, however, without some sources of perturbation—such as the 
increasing viscosity of the lava, the gradual blocking of the outlet by the addition 
to its walls of congealed rock, such as we see in hollow and other dykes, the 
formation of another outlet at alower level on the same radial rift, or the formation 
of anewone. The effect of such disturbing influences is illustrated in the case of 
the outpour of lava on and after the 27th March, 1905, where two lateral openings 
gave forth fluid rock, notwithstanding a difference in altitude of 65m. During 
the great eruption itself the same thing occurred ; lava issued at very considerably 
different levels. 
In the ordinary Vesuvian activity of a volcano, I repeat that two variations 
are constantly observable—those which for many years in my writings I have 
described as the lava-cake and cone-forming stage, and the dust and lapilli and 
crater-forming stages. The first is characterized by the emission of fragments of 
incandescent plastic scoria, or purely essential ejecta, the second by the expulsion 
of hard, cooled, broken-up fragments of the crater-walls, or purely accessory ejecta. 
The second is usually present when the ordinary Vesuvian grade of activity tends 
to pass to a paroxysmal stage. 
The first or cone-forming stage is due to the lava-level line being very high in 
the chimney, through and trom which the vapour bubbles out. In the act of 
TRANS. ROY. DUB, SOC., VOL. IX., PART VIII, 2C 
