May 30, 1884.] 
in general, that volcanic sand is composed 
chiefly of crystalline fragments, and contains 
a lower percentage of silica than the lava to 
which it belongs. With volcanic dust, how- 
ever, the case is very different. That which 
fell in Scandinavia, March 29 and 30, 1875, 
after having been carried by the wind from the 
great eruption in Iceland a distance of at least 
seven hundred and fifty miles, was composed 
almost exclusively of irregular, angular parti- 
cles of volcanic glass. Through the kindness 
of Professor Rosenbusch, in Heidelberg, I have 
obtained various samples of volcanic dusts for 
comparison. In the accompanying figures, 
series 1 represents the acute, angular, curved- 
sided fragments which are common in the 
Norwegian dust. In an excellent article by 
Murray and Renard, which appeared in Nature, 
April 17, 1884, p. 585, the forms of vitreous 
particles of Krakatoa dust are represented. 
It is undoubtedly true that the shapes. repre- 
sented are those which prevail in volcantfc dust, 
but they appear to be less characteristic than 
the curious outlines of fragments from the same 
dust given in series 2. In the succeeding 
series (3) are outlined the less common frag- 
ments in rhyolitic dust, collected by Mr. I. C. 
Russell along the Truckee River, in western 
Nevada. That these acute, angular, curved- 
sided forms are the most characteristic ones 
of volcanic glass particles, is impressed by a 
study of old tufas, in which the glass, origi- 
nally mixed with other clastic material, is 
completely replaced by quartz. An interest- 
ing tufa of this kind occurs at Breakheart Hill, 
Saugus, north of Boston. Where vitreous frag- 
ments of the most common shapes are re- 
placed by another material, the pseudomorph 
does not always suggest the original constitu- 
ent ; but when we find such forms as are rep- 
resented in series 4, from the Breakheart-Hill 
tufa, there can be but little question as to the 
original presence of volcanic glass. 
Krakatoa dust which fell at Batavia has 
been analyzed by Mr. Renard, and found to 
contain 65.04 % of silica, while the pumiceous 
form of the same lava, according to Mr. Id- 
dings, contains only 62 % of silica. It is well 
known that volcanic dust is composed chiefly 
and essentially of minute particles of natural 
glass; and, so far as definite observations 
have been made, they warrant the general as- 
sertion, that with occasional exceptions, which 
can be readily explained, volcanic dust con- 
tains a higher percentage of silica than the 
lava to which it belongs. 
Volcanic sand and dust must be regarded 
as differing, not merely in the size of their 
SCIENCE. 
653 
particles, but also in their physical and chem- 
ical constitution ; sand being composed, in the 
main, of crystalline fragments, and containing 
less silica than its corresponding lava, while 
volcanic dust is made up chiefly of glassy 
particles, which have a higher percentage of 
silica than the magma from which they were 
derived. Between these two extremes there 
are, of course, all possible intermediate terms ; 
but, nevertheless, it is evident, that, as a result 
of the operation of natural causes, there is a 
decided tendency, in connection with violent 
eruptions, to separate the magma into a basic 
and an acidic portion. ‘The degree of sepa- 
ration ultimately attained depends upon the 
final influence of the atmosphere upon their dis- 
tribution. Under favorable conditions, the dust 
may be spread many hundreds of miles from 
its source, while the sand is scattered within 
a comparatively small radius; but, under less 
violent and favorable conditions, both may be 
precipitated near the crater from which they 
issued. 
The inception of this divisional process is 
to be found in the condition of the magma be- 
fore its eruption. It is well known that crys- 
tals are frequently, and sometimes abundantly, 
developed in a magma ; so that, before its extru- 
sion, the magma may be regarded as made up 
of a crystalline, solid portion, and an amor- 
phous, more or less fluent portion. These. 
are generally thoroughly intermingled, but oc- 
casionally they are arranged, as in obsidians, 
in alternating bands; and they differ from 
each other in several important particulars, 
besides those already mentioned. The earliest 
products of crystallization are basic minerals, 
such as the ores of iron, hornblende, and mica ; 
and, as the process continues, the amorphous 
portion of the magma becomes more and more 
siliceous. On this account, the crystalline 
portion of the magma does not contain as 
high a percentage of silica as that which is 
amorphous. Capt. Dutton, in his researches 
upon the volcanoes of the Hawaiian Islands, 
made the interesting observation, that, at the 
moment a magma solidifies, a large quantity 
of vapor of water is given off. In the process 
of crystallization, the gases absorbed in the 
magma are rejected from the crystallizing sub- 
stances, and accumulate, under enormous ten- 
sion, in that portion which is amorphous. In 
this manner, the non-crystalline portion of 
the magma becomes stored with explosive 
compounds, under such stress, that, when the 
pressure is relieved, they may blow it to fine 
siliceous glass-dust ; while the crystalline, solid, 
basic portion of the magma, pulverized rather 
