530 Transactions. — Geology. 



occurs at any time between December and April. It must be 

 noted, too, that great submarine volcanic movements do occur 

 in this vast ocean no observations of which are made or records 

 kept, and, as I have said, the observations made are very con- 

 flicting. Sometimes the pumice-stone reaches the surface, 

 and for days vessels have sailed through a sea of it. But 

 more often, in the great depths, the outthrow remains where 

 it has been ejected, and is but slightly affected by ocean 

 currents. From one view of this intricate question a volcano 

 is really a delicate spring-balance. The boiling lava in the 

 great crater of Maun a Loa, at the Sandwich Islands, is kept 

 at its balance by so much sea-water finding its way beneath 

 the earth's solid crust by some crevice within a radius of per- 

 haps five hundred miles, more or less. An extra quantity of 

 water pouring in produces a lava overflow. If the whole 

 ocean-bed upon which the Sandwich Islands stand for a 

 radius of, say, five hundred miles gently crumpled or sub- 

 sided I do not think the eruptions would be so sudden. Still 

 it will be noted that the lava flowed in 1855 for thirteen 

 months. But even this is too short a period for the subsi- 

 dence theory as being the cause of volcanic action. Nor do 

 land-levels on the sea-line at the Sandwich Islands show sub- 

 sidence — at least, I have not heard or read of such. These 

 should show if subsidence was the cause of the many lava 

 outflows there. 



My task in this paper, I know, is to record the volcanic 

 phenomena of the Pacific. But it is evident that water pour- 

 ing in beneath the earth's crust — say, within a thousand miles 

 of the right or left of the 180th parallel of longitude — and being 

 immediately converted into steam, might tend to cause vol- 

 canic eruptions, and outputs of purely local lava, ash, or 

 cinder at different rents right along that parallel south of the 

 equator ; or the water might pour in immediately at the foot 

 of each particular volcano, which itself has formed the crevices 

 simply by upheaval. (I ask to be allowed to draw a distinc- 

 tion between a fault and a crevice.) In our own great Tara- 

 wera eruption of 1886 a small lake disappeared, I think, im- 

 mediately before the eruption. In the vast region of the 

 earth's surface now under consideration it may be that earth- 

 quakes breed earthquakes and volcanoes multiply volcanoes, 

 for it is evident that the more the earth's crust is broken 

 upwards the more readily will the sea-water find entrance 

 through the faults and crevices. On the great continents the 

 absence of active volcanoes may be accounted for by the fact 

 that only rain-water falls upon their surface, and that this 

 supply is not sufficient at any one fault or crevice to set up 

 thernial energy. Be this as it may, the fact of Mount Erebus, 

 Tarawera, Falcon Island, and Metis throwing out somewhat 



