Prof. J. W. Judd—The Lavas of Krakatoa. 5 
composition, at the other end one which is distinctly acid; while 
the others may be fairly classed as intermediate between those types. 
Yet all these rocks contain precisely the same mineral constituents, 
and the difference in their chemical composition is clearly due to the 
very different proportions in which their more basic constituents, 
the porphyritic crystals, are mingled with the acid material, the 
enveloping glass. 
Of course any variations in the relative proportions of the different 
felspars, pyroxenes and of magnetite will to some extent modify the 
ultimate chemical composition of the rocks; but I am convinced 
that in all these rocks the proportion of the several minerals to one 
another does not vary very greatly; in all of them felspar-crystals 
greatly predominate over the pyroxenes and magnetite; and among 
the pyroxenes the orthorhombic enstatite is almost always more 
abundant than the clinorhombice augite. 
Assuming, what is certainly not far from the truth, then, that the 
crystalline and glassy portions of these rocks have respectively the 
same composition in all of them, we can calculate from the data 
before us the relative proportions of the crystalline and glassy 
portions of each of these rocks. The result is as follows :— 
Glassy base Crystalline portion 
(parts in 100). (parts in 100). 
Most basic dykes at Santorin ...... 10 SSan00 90 
Other dykes at Santorin ............ 1O—35" tsenee 90—6d 
Buffalo Peaks lava ........0..00.000+ 33h asta 663 
Cheviot=nocksm manent cecsecee: GEN) Mikes eee 34 
Recent lavas of Santorin...:........ Sng ahis dase 50 
Lavas of Krakatoa .........ses.s0e+ SOR ig Lonasters 10 
It will be seen that at one end of the series we have the crystalline 
constituent nine times more abundant than the glass, and at the 
other end the glass nine times as abundant as the crystalline con- 
stituent. But in spite of this, the rocks according to the ordinary 
system of nomenclature, based on mineralogical constitution, must be 
called—and indeed have been called—by the same name, that of 
« hypersthene-andesite” or ‘“ enstatite-andesite.” 
Let us now turn our attention to another remarkable set of facts 
with respect to these lavas of Krakatoa. It is a very striking 
circumstance that all the materials ejected from the central vent of 
Krakatoa agree in a very marked manner in their chemical com- 
position and mineralogical constitution, both qualitatively and quanti- 
tatively. They all contain about 70 per cent. of silica, and are 
composed of 90 per cent. of base, with ten per cent. of porphyritic 
crystals ; these last consisting of plagioclase felspar, enstatite, augite 
and magnetite, always in nearly the same relative proportions. 
Nevertheless, while presenting this almost absolute identity in 
chemical and mineralogical constitution, we find among these pro- 
ducts of Krakatoa three distinct types of materials—exhibiting the 
most striking differences in their physical characters and in the 
mode of their behaviour during ejection. These differences are seen 
to be dependent on peculiarities presented by the base or ground- 
