oj Edinhurgli, Session 1883 - 84 . 
479 
clieajical analysis, Ave pounded in an agate mortar several varieties 
of pumice, and the powder thus produced clearly showed itself to 
be composed of particles in Avhich were recognisable, with little 
trouble, the characters of the pumice-like material which is constantly 
met with in the sediments, and of which the ashes of Krakatoa give 
us beautiful examples. The diagnostic character to which we here 
make allusion rests on the distinctive peculiarities of incoherent 
volcanic products. What distinguishes them from lavas is not 
merely the extraordinary abundance of vitreous matters, but also 
the prodigious number of gas-bubbles which are enclosed by the 
pumice and vitreous volcanic sands and ashes. These bubbles 
are due to the expansion of the gases dissolved in the magma, 
Avhich also determine the eruption. If we admit, as everything seems 
to show, that these incoherent volcanic matters are the products of 
the pulverisation of a fluid magma, Ave can understand that these 
particles, on cooling rapidly, will remain in the vitreous state, and, on 
the other hand, that the dissolved gases, yielding to the expansion, 
Avill form numerous pores which will become elongated owing to the 
mode of projection. It is the existence of these bubbles, or of such 
a filamentous structure, AAdiich points out to us the vitreous volcanic 
materials in spite of the great fineness of subdivision. It is also 
this structure which allows these bodies to be carried to such great 
distances from the scene of eruption. 
The examination of the Krakatoa ashes, and of the dust resulting 
from the pulverisation of the pumice of that volcano, shoAVs 
markedly the peculiarity due to the bullous structure. If this 
grey-green pulverulent matter be placed under the microscope it is 
seen to be composed of almost impalpable grains, with a mean 
diameter of 04 mm., Avhich are almost exclusively colourless or 
broAvnish vitreous particles permeated by bubbles. The bubbles 
are rarely globular, but often elongated, as we have just pointed 
out, and they give a draAvn-out appearance to the fragments. As 
often happens, several bubbles are elongated parallel to each other, 
and in this case, the pore becomes a simple streak ; the fragment 
then assumes a fibrous texture, Avhich may cause it to resemble at 
first sight a striated felsiDar or an organic remnant ; but an exa- 
mination of the outline Avill never allow of this confusion. If Ave 
examine the terminal contours and lines of these bubble-containing 
fragments, we never find that they are straight lines, but that 
