HEAT VOLCANOES. 693 



nally, while usually cellular within. Such masses are called volcanic bombs. 

 They may have lenticular as well as spheroidal shapes. 



2. Movements of the lavas in the crater. — (a.) Upward move- 

 ment — As the vaporizable substances (water, sulphur, etc.) and at- 

 mospheric air expand while rising in the volcanic vent, they displace 

 correspondingly the lava, and so cause a general expansion of the 

 mass. This alone produces a rise of the lavas in the conduit. 



When the boiling of a viscid fluid in a tube causes its upper surface to 

 ascend, because the liquid at top becomes inflated or frothy with vapor, it 

 exemplifies the same principle, although the degree of inflation very far exceeds 

 that in a dense lava. The fact of a rising in the volcano from this cause is 

 beyond question. 



This rising becomes apparent in overflowings from the pools of 

 the crater, over its bottom, in streams which cool and become solid 

 lava. Whether the Avhole rising is due to this cause is not ascer- 

 tained. The risings and overflowings are repeated from time to 

 time, until the material within the crater has reached a height and 

 an intensity of action that lead to an eruption. 



At Kilauea (the bottom of which, when at its lowest mark, is 3000 feet above 

 the sea) the conduit of liquid lava descending downward below the crater is 

 3000 feet long to the sea-level ; and it may extend many miles, or perhaps scores 

 of miles, below this. Nineteen miles would correspond to about 100,000 feet. 

 A rise of the lavas within the crater of 400 to 500 feet in the manner above ex- 

 plained is all that in three cases of eruption at Kilauea preceded the outbreak. 

 Five hundred feet in 100,000 is an average expansion of only a half of one per 

 cent. But it is probable that the vapors which produce this result are com- 

 paratively superficial; they may be from the fresh or salt waters of the sur- 

 rounding region. 



The increase of activity as the lavas rise in a crater has two 

 obvious causes : (1) the temperature of the lavas increases with the 

 pressure ; and, consequently, a rise of 100 feet would have increased 

 very much the temperature at the bottom of that 100 feet, and so 

 on for greater depths ; (2) the rise exposes a higher column of 

 liquid lava above to the action of external waters. 



(b.) Circulating movement. — In the lava-conduit the greatest heat is 

 along the centre, most remote from the cold sides. Hence, as in 

 any cauldron, the ascent from inflation by rising vapors would be 

 greatest at the centre ; there would therefore be at the surface a 

 flow from the centre to the sides, and a system of circulation. This 

 was exhibited on a grand scale at Kilauea in 1840, where the liquid 

 lava in the great lake (1000 feet across, a, fig. 969) seemed like a 

 river that came to the surface for a moment and then disappeared. 



