48 FOYE. 



dike intruding the limestone is 20 feet in width. It is deeply laterit- 

 ized and outcrops in a long, well-defined depression in the limestone 

 which forms a perpendicular wall on either side, indicating that the 

 volcanic rock here weathers more rapidly than the sedimentary. 



On the northeastern side of Andivathi, ash-beds are seen to overlie 

 the limestone, and to dip southeast towards the eroded and nearly 

 vertical walls of limestone, lying just across a strait, not more than a 

 hundred yards wide. 



Again, at Mba Vatu, on the northeastern side of Vanua Mbalavu, 

 thin dikes, about a half inch in thickness, were found intruding the 

 elevated limestone at a height of 300 feet. 



Near the center of the large curve of the question mark, a peak, 

 (Koro Mbasanga), 920 feet high, was built up during the second 

 volcanic period and the cone of which it forms a part has since been 

 maturely eroded. This cone has been placed in the second period 

 of volcanic activity, since (1) its lavas are of the basaltic type and 

 unlike the lavas of the first or andesitic period; and (2) lavas traced 

 from this center overlie foraminiferal limestone which rests uncon- 

 formably on the andesites of the first period. 



Scattered over the surface of the deeply lateritized hills of andesite 

 are numerous silicified fragments of coral. It was noticed that the 

 surface waters, carrying silica from the lateritized hills, were apt to 

 deposit the silica near the contact of the limestone with the volcanic 

 rock and it is supposed the silicified coral originated in this way. The 

 presence of the corals indicates that considerable areas of the lime- 

 stone have been swept away. Whether the andesitic hills of Vanua 

 Mbalavu were ever wholly covered with limestone is not known, but 

 since the elevated limestone now occurs at elevations as great as that 

 of the highest andesitic hills, this is considered possible. 



The coastal irregularities of the old andesitic hills, as well as the 

 deep bays of the limestone portion of the island, have been appealed 

 to as evidence of recent submergence. It does not appear to the 

 writer that this evidence is necessarily conclusive. In the first place, 

 the valleys cut in the andesite may have been formed before subsi- 

 dence and, after the subsequent uplift, uncovered by the erosion of 

 the unconformably overlying limestones. In the second place, the 

 drainage passing from the andesitic hills ])en('ath the limestone or 

 percolating through the limestone to the andesitic contact, has pro- 

 duced effects peculiar to islands of this type. ]Much of the drainage 

 passes by deep channels out to sea and appears, off-shore, as fresh- 

 water springs. The persistence of these channels has led to the 



