650 GEOLOGY OP SAMOA AND THE ERUPTIONS IN SAVAII, 



and previous configuration of the country. I was informed by 

 reliable authorities, like Mr. Williams and the Rev. Pere Mennel 

 of Safotu, that originally a deep valley reached almost to the 

 present site of the active cone. The earliest flows followed this 

 valley, and in it, for a distance of several miles, there is now a 

 thickness of lava which probably exceeds a thousand feet. What 

 was originally a deep valley, only a few hundred feet high, now 

 forms a huge, bulging, lava ridge about 1500 feet high near the 

 volcano, and sloping gently towards the sea. This inclined lava- 

 plain is continually rising through the intercalation of new lava 

 sheets or sills under the cooled crust. Captain Allan, of the 

 s.s. " Maori," who periodically visits the islands, tells me that he 

 sees an appreciable difference in its altitude each trip. 



Most of the lava at present exuded goes to the sea, following 

 definite lava tunnels beneath the cooled crust. The course of 

 these tunnels can be traced by ventholes, through which steam 

 and sulphurous gases rise (see Plate lviii.). 



Where the lava flows into the sea we have what seems a 

 miniature volcano (Plates lix.-lx.). On the shore-side an embank- 

 ment of stones and cinders is formed by explosions, which 

 occurred at the time of my visits in very rapid succession (one 

 every two minutes), and which depend on the principle of geyser- 

 action. Immense masses of steam form, which, as shown in 

 Plates lix.-lx., rise like gigantic columns, and drift off as cumulus 

 clouds, giving continued rains on the area over which they drift. 

 The showers from these clouds often were salty and contained 

 cinders, showing that matter was carried up mechanically by the 

 steam. The roar of the rising columns was continuous, modified 

 only by the geyser-like explosions. In the vicinity waterspouts 

 and tornado-like whirlwinds were frequently formed. At night 

 this spot had a glow overspreading it like the volcano itself; and 

 on each explosion followed a magnificent firework display, 

 probably through red-hot stones being hurled up several hundred 

 yards. 



The cause of the explosions seems to be as follows. As the 

 lava pours into deep water, a considerable amount of water 



