miles in circumference at its base, has probably been produced wholly by its own eruptions. A large part of the chain of the Andes consists of volcanic rock, but the proportion we have not the means of estimating. 2. _Tertiary Lavas._--There is another class of volcanic products, which are so situated with reference to the tertiary strata that they must be referred to that period. The principal localities of these lavas, so far as yet known, are Italy, Spain, Central France, Hungary, and Germany. They are also found in South America. Those of Central France have been studied with the most care. They occur in several groups, but they were the seats of volcanic activity during the same epoch, and formed parts of one extensive volcanic region. Each of these minor areas, embracing a circle of twenty or thirty miles in diameter, is covered with hills two or three thousand feet in height, which are composed entirely of volcanic products, like the cone of Ætna. On many of them there are perfectly-formed craters still remaining. Numerous streams of lava have flowed from these craters, some of which can now be traced, throughout their whole extent, with as much certainty as if they were eruptions of the present century. Some of the lavas have accumulated around the orifices of eruption, forming rounded, dome-shaped eminences. These lavas generally consist of trachyte, and have therefore a low specific gravity, and imperfect fluidity. The basaltic lavas have often spread out over broad areas, and, when they have been confined in valleys, have reached a distance of fifteen miles or more from their source. There still remain indications of a current of lava which was thirty miles long, six broad, and in a part of its course from four to six hundred feet deep. The above sketch (Fig. 5) will give some idea of the highly volcanic aspect which the district of Auvergne, in France, presents. [Illustration: Fig. 5.] The unimpaired state of some of the cones and craters, and of the lava currents, would lead to the impression that these regions have been the theatre of intense volcanic action within a very recent period. But there is good reason to believe that this has not been the case. "The high antiquity of the most modern of these volcanoes is indeed sufficiently obvious. Had any of them been in a state of activity in the age of Julius Cæsar, that general, who encamped upon the plains of Auvergne and laid siege to its principal city, could hardly have failed to notice them."