molecular gaseous expansion and not by the crushing of old rocks nor by the collapse of crater walls. This forms a striking example of the process of ash making as described in a paper by the present writer entitled "Volcanic vortex rings and the direct formation of ash from liquid lava" and which appears in the November issue of the American Journal of Science. During this eruption the explosions, although very powerful, were not sufficiently sharp to produce the phenomenon of the "flashing arcs" as observed by me at Vesuvius (1906) and Etna (1910) and described in the American Journal of Science.

In addition to the solid blocks and the ash, a very large quantity of porous, vitreous scoriæ or "lapilli" were ejected. These were of the same material as the ash and containing in some cases inclusions of already formed crystals of angite. In falling, they were still plastic to the extent of conforming to the forked branches of the Genesta plants and even to be impaled upon the spines (pl. 6).

This eruption was very instructive. A careful study was made of the great "mushroom vortices" (pl. 7, fig. 1), by whose mechanism heavy rocks are carried to a great height and then thrown to a distance by the vortex whirl.

Some idea of the crater, 300 meters in diameter, of this eruption may be had from plate 7, figure 2.

On several occasions, including the present one, the writer has been completely enveloped for 15 or more minutes at a time in the cloud of gas and ash proceeding directly from the crater of a volcano during a paroxysmal eruption. In every case there was no noxious gas-no HCI, SO2, H2S, CO,-in preceptible amounts, although these are present in distressing quantities during phases of minor activity. The conclusion is inevitable that the paroxysmal gases-i. e., the gases which produce a great eruption and which have been the cause of the formation of volcanoes-consist mainly of the same ingredients as atmospheric air. Under the above mentioned circumstances I have found only a slight feeling of oppression, which may be due to the high temperature or possibly to a slight deficiency in the proportion of oxygen due to oxydations during the subterranean travel of the gases.

These two eruptions prove that Stromboli shares in the general increase of activity of the Italian volcanoes.

Mount Etna is preparing for a great eruption.

DATA OF THE ERUPTIONS 1907 AND 1912.

1. The ash of 1907 was acid, that of 1912 alkaline, both noxious to green vegetation.

2. The paroxysmal crises of both eruptions corresponded with the luni-solar phases, to which Stromboli is very sensitive.

3. The ash emission of the 1912 eruption was accompanied by strong electrical manifestations, the volcanic lightnings being vivid and almost incessant.

4. The products of the fumaroles of this eruption, collected by the writer and analyzed by Dr. Henze, of the Naples Aquarium, consist principally of Al, Fe, and Mg, combined with H2SO, and HCl. Their gaseous emanation, as tested on the spot, was chiefly SO, and a trace of H,S.

5. The eruptive months are six in number and have been designated by the writer with the letters A to F, inclusive.

6. The eruption proper lasted from July 22 to August 14, 1912. 7. The crater of 1907 was 200 meters in diameter, very deep, and the throat of the volcano was left quite free, permitting the gases to escape quite continuously after the eruption.

8. The eruption of 1912 enlarged the former crater to 300 meters and left it partly filled with collapsed cone material. This tends to confine the gases until they break through at intervals and form an ash cloud, giving the impression of a greater activity than really exists. 9. The paroxysmal gases, as at Vesuvius in 1906, had approximately the composition of atmospheric air.