BY H. I. JENSEN. 663 



a long fault caused by an overfolding movement from the west- 

 north-west, which is tending to elevate the sea-bottom along its 

 whole length; and the deep water to the west of it is accounted 

 for by a downthrow to the rear of the fault. The breaks in the 

 continuity in the shelf between the volcanic line and the deep 

 trough eastward, may be due to cross faulting of the blatter 

 type, and its after-result cross trough-faulting.* The fact that 

 tourmaline inclusions have been discovered in the Eua lavas, and 

 the existence of a deep trough east of Tonga point to these 

 islands being situated over continental rocks on an old shore-line. 

 The fact that the great deep is separated by a ridge from the 

 fissure-line shows that it is probably a structural feature, and not 

 the effect of volcanism. The simultaneity of eruptions along the 

 line of fissure can be accounted for only on the assumption that 

 crustal movements take place simultaneously along the whole 

 length of the fissure, and the extravasation of lava is probably 

 due to the melting of rock by pressure and friction below. 



7. Connection op the Eruption with Lunar Phases, 

 Sunspots and Earth-Magnetism. 



It has been noticed by the observers of the volcano in Savaii 

 that most of the important eruptions, or periods of particular 

 violence, have been at New Moon, or early in the first quarter; 

 some occurred at Full Moon, and very few in the last quarter. 



Dr. Linke, of the Apia Observatory, informs me that very great 

 magnetic disturbances were recorded by his instruments at the 

 time of the outbreak of the volcano in 1905. Every subsequent 

 violent eruption was likewise accompanied by magnetic dis- 

 turbances How far these disturbances depend on solar or other 

 extra-terrestrial causes, or on the movement of lava-masses, rich 

 in ferric minerals, within the earth, is a matter of extreme 

 importance in the realm of seismic forecasting; and we may look 

 to Dr. Linke for the solution of the problem, as regards Savaii. 



There seem to be good reasons to suppose that great earth- 

 movements or lava-movements under an area will affect local 



* Gf. "Geology of the New Hebrides," by D. Mawson. These Proceedings, 

 1905, Part 3. 



