GEOLOGICAL OBSERVATIONS I\ FIJI. 39 



From the central ridge steep slopes, of 30 to 35 degrees, descend to 

 the sea. The slopes of the central and southern parts of the island 

 are largely underlain by basaltic flows and agglomerate. It is difHcult 

 to reconcile the steepness of the slopes with the general fluidity of 

 basaltic lava, if basalts constitute the whole depth of the ridge down 

 to the sea-level. The problem becomes less serious if it be assumed 

 that the basalts actually form a relatively small part of the mass; 

 that is, a veneer on the older andesites exposed at the northern end of 

 the island. In certain places, the basalts present an abrupt change of 

 le^'el as they approach the coast, as though they had flowed out over 

 the bench on the older andesites above described. 



At the southern end of the island, a series of seven or eight cones, 

 700 to 800 feet in height, rise above a nearly flat basaltic plain. These 

 cones are arranged along a north-south line, approximately parallel 

 to the main ridge of the island. 



Two different explanations of the isolated reef at the southern end 

 of Taviuni are conceivable. It may represent part of an old reef 

 which was not entirely covered by the recent basalt flows. If so, the 

 depths of 50 to 75 fathoms close to the shore represent the original 

 descent from the edge of the reef into the ocean abyss. On the other 

 hand, the existing reef may be due to the colonization of the flat sur- 

 face of a recent lava flow extending beneath the sea. The latter 

 hypothesis leaves unexplained the considerable depths close to the 

 land. Block-faulting may account for the sharp descent. 



It is easy to postulate faulting in order to explain certain features 

 about Taviuni but difficult to prove it. For example, Goat Island, 

 lying two or three miles off the west coast, is composed of ash-beds 

 which are gently folded and dip towards the basaltic slopes of Taviuni. 

 The present dip would seem to indicate that the source of this ash 

 lay to the west. A deep channel, however, separates the island from 

 Vanua Levu in that direction. It is not impossible that faulting has 

 formed the strait and destroyed the center of eruption from which the 

 ash was ejected. Yet the beds may simply be an erosion remnant cut 

 off from Taviuni. 



Again the small island, Vunimbani (200 feet high), off the northeast 

 coast of Taviuni is capped by 15 to 20 feet of water-laid ash, dipping 

 5° N. E., and overlying flows and agglomerates. The island is con- 

 nected with Taviuni by a flat exposed at low tide. The coast of the 

 larger island at this point is composed of deeply lateritized basalts 

 whose topography is quite unlike that of Vunimbani. 



The accompanying sketch (Figure 19) was made from the north- 



