738 DYNAMICAL GEOLOGY. 



(2.) Catastrophic Eruptions. — In other cases, especially when the 

 lavas are quite viscid, as at Vesuvius, where the eruptions are usually 

 attended by earthquakes, they depend perhaps most largely on the 

 access of marine or fresh waters, and the consequent sudden produc- 

 tion of vapors ; but even in such, there is usually a gradual increase 

 in height of the lavas going on through years, preparing for the catas- 

 trophe. 



(3.) Projectile Power inversely as the size of the Crater. — Great 

 craters, like the Hawaiian, have little projectile power because of the 

 liquidity of the lavas ; and little ones large, for the reverse reason. 

 Free liquidity leads to great size of the crater, as well as a pit-like 

 form ; and viscidity to relatively narrow throats. With the former, 

 cinder ejections are usually few and low and the cones that are made 

 are small ; with the latter, large and high, because the vapors are 

 kept from escaping until they have collected into very large bubbles, 

 and the cones are sometimes several thousands of feet high. 



(4.) Periodicity in Volcanic Phenomena. — The crater of Kilauea 

 took eight years for filling 400 to 500 feet of its depth preparatory to 

 the eruptions of 1832 and 1840; and it was as full again in another 

 eight years ; but, this time, owing to a decline of activity, for some 

 reason, an eruption did not occur. In the case of the summit-crater of 

 Mount Loa, three great eruptions have taken place with intervals of 

 3i years, and six in the course of 25 years. 



(5.) Cinder Cones rapid, in formation. — The ejections of cinders 

 are often very copious, and the cone, or several along the line of a 

 fissure, may grow to a height of hundreds of feet in a few days. 



(6.) Eruptions of Lavas usually through Fissures. — Volcanic moun- 

 tains are not often strong enough to stand the pressure of liquid lavas 

 carried up to the very brink of the crater. Hence, a volcanic moun- 

 tain is intersected by great numbers- of dikes. Mount Loa, although 

 twenty miles in breadth only ] ,800 feet below its top, has now no 

 summit overflows, but becomes fractured by the eruptive forces, and 

 so discharges its lava-floods. And usually the whole course of a stream 

 of lava is the course of a supplying fissure. It is probable that Kil- 

 auea began in the opening of a great fissure at a Mount Loa eruption. 



Owing to eruptions through fissures, tufa and cinder cones often 

 occur in lines over a volcanic mountain. A fissure may continue its 

 ejections at its widest parts long after its general outflow has ceased, 

 and end in throwing up cinders. But the fissure, and not the subor- 

 dinate cones, gives out the chief part of the ejected lava. 



It is a remarkable fact, shown by the relative amount of degradation in the moun- 

 tains, that the western end of the Hawaiian series of islands has been longest extinct; 

 and so also the western or northwestern volcano of several of the islands, as in the case 



