56 G. F. Becker—Texture of Massive Rocks. 
granular portions, and the constitution of the micaceous por- 
tions must be different from that of either of the others. In 
‘short, original differences in composition are necessary and 
sufficient to explain the differences in texture; and since the 
composition varied enough to produce striking differences in 
texture, the magma can never have been homogeneous and 
therefore it can also never have been thoroughly fluid. It also 
seems quite clear, from the results reached in preceding para- 
graphs, that those portions which were more fusible and became 
fluid, solidified as porphyry, while those portions which were 
only reduced to a pasty condition and which consisted of rather 
intimate mixtures of somewhat heterogeneous material, solidi- 
fied as a granular rock. The melting point of most bodies 
rises with the pressure.* It is therefore in so far conceivable 
that a column of homogeneous eruptive rock of uniform tem- 
perature should exist, the upper portion of which is fluid and 
the lower portion pasty. Such a column might solidify to a 
porphyry near the upper end and to a granitic mass at the lower 
end. ‘The lack of knowledge of the “relation between melting 
points and pressures of minerals makes it impossible to assert 
that such cases may not sometimes occur in nature, but, if so, 
they must be exceptional. Positive evidence has been given 
above that porphyries such as andesites must be fluid enough 
to allow of free crystallization before eruption and at pres- 
sures which 1 beheve must greatly exceed those at which any 
rocks accessible to dwellers on the surface of the earth can 
have solidified. A column of andesite of which the lower por- 
tion was rendered pasty by mere pressure would have to exceed 
in length the distance from its point of eruption to the source 
of the lava, probably not less than five miles and very likely 
much more. Again if the lowerend of a column, say 10,000 
feet in length, were rendered pasty by mere pressure, it would 
* Lavas not seldom contain crystals of primary consolidation which appear to 
be rounded, as if they had been partially fused after crystallization. The forma- 
tion of black borders on hornblende may also indicate a partial fusion or soften- 
ing of these crystals. These phenomena are almost certainly attendant upon 
eruption, and if they indicate real fusion, lead to the conclusion that the melt- 
ing point of such minerals actually falls with diminishing pressure. Bodies, the 
melting point of which varies in the same sense as the pressure, are denser in the 
solid than in the fluid state at the melting point, and this question consequently 
has a bearing on the problem of the solidity of the earth. In the Hawaiian lavas 
the glass ejected has substantially the composition of basalt and the fact that, as 
Mr. King observed, the chief constituent minerals sink in and are heavier than 
this glass, points in the same direction; though unless there are really indications 
of fusion of the crystals it remains uncertaim how far from this melting point the 
separation by density took place. Other purely geological phenomena also tend 
to the same conclusion. Thus Captain Dutton has observed that the crusts which 
form on the melted lavas in the craters of Hawaiian voleanoes often break up and 
sink. These crusts are of course largely glass, but glass is certainly as a rule 
less dense than the minerals which crystallize from it, so that were the crusts 
holocrystalline they would sink ouly the more readily. 
