8 Prof. J. W. Judd— The Lavas of Krakatoa. 
Those who regard the bulk-analysis of an igneous rock as 
necessarily representing the composition of the magma out of which 
the minerals which compose it have separated, and who further 
consider the successive crops of crystals formed in a rock as being 
determined by the alteration in the composition of the residual 
magma by the constant separation of the basic minerals from, it will | 
find it difficult to harmonize such views with the facts presented to 
us at Krakatoa. 
It is difficult, in the first instance, to account for the separation of 
precisely the same minerals, the felspars, pyroxenes and magnetite, 
in magmas of such diverse composition as is represented in the 
third table on page 4. On the other hand, it is a significant circum- 
stance that the glass of the Krakatoa-lavas is sometimes found 
without the porphyritic crystals; pumice formed of this glass 
without any enclosed crystals was certainly erupted during the 
eruption in May, 1883. Similar facts are recorded by Fouqué in 
the case of the Santorin lavas. It must be remembered, too, that the 
whole of the lavas of Krakatoa, and perhaps of all the volcanoes on 
the line of cross fissure upon which it is situated, are of an exceptional 
character; the ordinary enstatite-andesites of Java containing only 
about 55 to 62 per cent. of silica, and having a much smaller pro- 
portion of glass to the porphyritic crystals. In Santorin, too, we 
have the same difference, though somewhat less marked, between 
some of the older and deep-seated dykes and the modern lavas of 
the volcano. 
All of these facts point to the conclusion that after the partial 
separation of a magma into crystals and a colloid residue, the two may 
be separated by a process of liquation, and subsequently become 
mingled again in varying proportions. I need scarcely point out 
that the condition of many porphyritic crystals—eroded, broken, or 
frayed out in the magma in which they lie--are strongly suggestive 
of the same conclusion. These porphyritic crystals indeed have 
often the appearance of foreign materials which have been caught 
up by the magma in which they are now found. 
It is a significant circumstance that the Krakatoa-rocks, and this 
is true of many other rocks of the same class, do not appear to have 
their porphyritic crystals scattered quite at hazard through their 
mass. On the contrary, the felspars, pyroxenes and magnetite are 
seen to form little knots often intercrystallized in the midst of the 
vitreous base. In this respect, there is an approach to the structure 
for which I have proposed the name of “ glomero-porphyritic,” in 
which we find what looks like the association of minerals typical of 
one class of rock scattered through a rock of a totally different class. 
In the typical example of the kind which I have described fragments 
of troctolite appear as if scattered through an ophitic dolerite. 
In the case of the plutonic rocks of the Western Isles of Scotland, 
I have shown how frequently one or other of the constituents of a 
magma may segregate iocally so as to give rise to a variety of rock 
in which this mineral predominates. In this way gabbros are found 
graduating in places into rocks essentially composed of olivine, 
pyroxene, or felspar. 
