of Edinburgh, Session 1883-84. 
487 
them easily under the microscope. We would remark, however, 
that the presence of crystals, either of hypersthene, of augite, or of 
particles of magnetite in an atmospheric dust collected in Europe, 
does not prove in a certain manner that the dust belongs to the 
ashes from Krakatoa; for besides the difficulties of an exact minera- 
logical determination of the fragmentary elements, it is difficult to 
understand how these heavy minerals should have been carried by 
the aerial currents, while the vitreous dust is absent. As we have 
just shown, it is the contrary which should have taken place. 
It results as a corollary from these considerations that the chemical 
composition of an ash may vary according to the point at which it 
has been collected, and it tends also, other things being equal, to 
become more acid the further it is removed from the centre of 
eruption. If we admit, for example, that the magma which gave 
birth to the ashes of Krakatoa is an augite-andesite, as everything 
seems to indicate, the percentage of silica (65 per cent.) which our 
analysis shows appears too high, but if we remember, what we have 
just said, that the ashes become deprived, during their passage 
through the atmosphere, of the heavier and more basic elements, 
it will be understood that the vitreous and felspathic materials, 
which have a lower specific gravity, and are, at the same time, 
more acid, will accumulate at points farthest from the volcano. It 
will be sufficient to have directed the attention to this fact to show 
how the percentage of silica in the ashes from the same eruption 
may vary according as they are collected at a variable distance from 
the crater. 
The predominance of vitreous splinters in deep sea sediments far 
removed from coasts is even more pronounced than in volcanic ashes 
collected on land. This arises, as we indicated at the commence- 
ment, from the large quantity of pumice carried or projected into 
the ocean, whose trituration, which takes place so easily, gives 
origin to vitreous fragments difficult to distinguish from those 
projected from a volcano in the form of impalpable dust. In 
addition, we may state that in the distribution of volcanic materials 
on the bottom of the sea, the ashes are subjected to a mode of 
sorting having some analogy to that which takes place during 
transport through the atmosphere. When these ashes fall into 
the sea a separation takes place in the water; the heaviest particles 
