to the west.

The total mass ejected has been estimated at 60 milliards of square yards.

Stromboli, in the Mediterranean (Fig. 23), though only 2500 feet in height, is imposing from its superb regularity, and its roots plunge below the surface to a depth of 4000 feet.

It is, moreover, very interesting from the regularity of its action, which has a period of 5 minutes or a little less. On looking down into the crater one sees at a depth of say 300 feet a seething mass of red-hot lava; this gradually rises, and then explodes, throwing up a cloud of vapour and stones, after which it sinks again. So regular is it that the Volcano has been compared to a "flashing" lighthouse, and this wonderful process has been going on for ages.

Though long extinct, volcanoes once existed in the British Isles; Arthur's Seat, near Edinburgh, for instance, appears to be the funnel of a small volcano, belonging to the Carboniferous period.

The summit of a volcanic mountain is sometimes entirely blown away. Between my two first visits to Vesuvius 200 feet of the mountain had thus disappeared. Vesuvius itself

R

stands in amore ancient crater, part of which still remains, and is now known as Somma, the greater portion having disappeared in the great eruption of 79, when the mountain, waking from its long sleep, destroyed Herculaneum and Pompeii.

As regards the origin of Volcanoes there have been two main theories. Impressed by the magnitude and grandeur of the phenomena, enhanced as they are by their destructive character, many have been disposed to regard the craters of Volcanoes as gigantic chimneys, passing right through the solid crust of the globe, and communicating with a central fire. Recent researches, however, have indicated that, grand and imposing as they are, Volcanoes must yet be regarded as due mainly to local and superficial causes.

A glance at the map shows that volcanoes are almost always situated on, or near, the sea coast. From the interior of continents they are almost entirely wanting. The number of active volcanoes in the Andes, contrasted with their absence in the Alps and Ourals, the Himalayas, and Central Asian chains, is very striking. Indeed, the Pacific Ocean is encircled, as Ritter

has pointed out, by a ring of fire. Beginning with New Zealand, we have the Volcanoes of Tongariro, Whakaii, etc.; thence the circle passes through the Fiji Islands, Solomon Islands, New Guinea, Timor, Flores, Sumbava, Lombock, Java, Sumatra, the Philippines, Japan, the Aleutian Islands, along the Rocky Mountains, Mexico, Peru, and Chili, to Tierra del Fuego, and, in the far south, to the two great Volcanoes of Erebus and Terror on Victoria Land. The Atlantic is in this respect, as in some others, a great contrast to the Pacific.

We know that the contraction of the Earth's surface with the strains and fractures, the compression and folds, which must inevitably result, is still in operation, and must give rise to areas of high temperature, and consequently to Volcanoes. We must also

remember that the real mountain chains of our earth are the continents, compared to which even the Alps and Andes are mere wrinkles. It is along the lines of the great mountain chains, that is to say, along the main coast lines, rather than in the centres of the continents, which may be regarded as com

paratively quiescent, that we should naturally expect to find the districts of greatest heat, and this is perhaps why Volcanoes are generally distributed along the coast lines.

Another reason for regarding Volcanoes as local phenomena is that many even of those comparatively near one another act quite independently. This is so with Kilauea and Mouna Loa, both on the small island of Hawaii.

Again, if Volcanoes were in connection. with a great central sea of fire, the eruptions must follow the same laws as regulate the tides. This, however, is not the case. There are indeed indications of the existence of slight tides in the molten lake which underlies Vesuvius, and during the eruption. of 1865 there was increased activity twice a day, as we should expect to find in any great fluid reservoir, but very different indeed from what must have been the case if the mountain was in connection with a central ocean of molten matter.

Indeed, unless the "crust" of our earth was of great thickness we should be subject to perpetual earthquakes. No doubt these are