314 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1909. 
As we go deeper in the volcanic.conduit these same differences will 
be exaggerated, so that as the magma escapes almost all the glass 
may be converted into microliths so as to leave little material to hold 
them together, whilst the evolution of volatile gases will be sufficient 
to separate them into dust, producing the pulverized material con- 
stituting the essential part of the last and topmost deposit -of an 
explosive eruption. 
If the igneous paste rises from still greater depths it may come 
from little or nonaquiferous parts of the conduit, which, together with 
a higher specific heat from fewer losses thereof, will allow it to gush 
forth in a nonfragmented state as a lava. 
The want of uniformity in the water-bearing rocks at different 
depths is too well recognized for us not to see the influence on pos- 
sible departures and irregularities which might result in the sequence 
of ejecta having the characters I have shown you as the simplest ex- 
pression of an eruptive phase. 
In an open chimney of a volcano in chronic activity the constant 
circulation up the voleanic conduit allows of too little time for the 
magma to acquire much volatile materials, and it is only when this 
outflow is more or less impeded at the vent that more volatile ma- 
terials are acquired and the volcano assumes paroxysmal or explo- 
sive fits. 
In conclusion, I may say I have tried to summarize the trend of my 
researches for the last thirty years, and if you will try to read the 
whole phases of voleanic activity in this ight you will find it the only 
satisfactory explanation universally applicable to all cases of the 
eruptive mechanism. No other theory that has been advanced has 
ever been based on the characters of the actual essential ejecta, and no 
other one fits without exception the whole range of the very varied 
phenomena of volcanicity. 
EXPLANATION OF PLATES 1, 2, AND 3. 
PLATE 1. 
A Shells of contraction. 
B Neutral zone or zone of neither contraction nor compression. 
C Dry shells of compression. 
D Aquiferous shells of compression. 
BE Fissure between shells of maximum cooling and contraction filled by lique- 
faction of the edges of these shells by diminished pressure. 
F Same, but in the shells of less cooling and contraction. 
G Fissure extending up between two areas of compression and islands of con- 
traction, but not reaching the aquiferous shells. 
H The same, but reaching into the aquiferous rocks. 
I The same, but having reached aquiferous shells, has been enabled to extend 
upward by explosive action into a laccolite and sill in one case and directly 
to a voleano in the other. ; 
