140 FOYE. 



hypersthene andesite, hornblende andesites, and rhyolites; from the 

 Lau Islands, augite andesites and basalts; and from Kandavu, basalts, 

 augite and hornblende andesites. 



The products of the later, basaltic period are not so widespread as 

 those of the previous periods. Basalts are known from Viti Levu; 

 olivine, augite and hornblende basalts from Mbengha; basalts from 

 the Yasawa islands; basalts from Vanua Levu; and basalts from the 

 Lau group. 



In view of the diversity of lavas assigned to the second andesitic 

 period, it is important to note that the two most recent eruptions 

 known in Fiji have a very diverse composition. The recent cones 

 of Taviuni gave vent to basaltic flows, whereas Mount Washington, 

 Kandavu, is a hornblende-andesite volcano. 



Differentiation in the Lavas of Fiji. 



The discussion of the volcanic succession has brought out an im- 

 portant relationship between the rocks of Fiji. In the islands of the 

 Lau Group, which appear to be of recent origin, there is little diversity 

 in the lavas of the two known periods. Basalts, or augite andesites 

 related to basalts, characterize both periods. 



On the other hand, within the larger islands, the diversity of lava 

 is great. It is pertinent to inquire, why hornblende is so frequently 

 found in the rocks of Kandavu, Mbengha, Viti Levu and Vanua Levu 

 and is not known on the rocks of the Lau islands. 



In most cases, the hornblende gives evidence of being resorbed into 

 the magma and of breaking down into augite and magnetite. It 

 apparently was not in equilibrium with the magma at the time of its 

 solidification. If it is true that hornblende and biotite form only in 

 the presence of magmatic gases, especially water, the loss of these 

 gases on extrusion may be considered to be the cause of the instability 

 of the hornblende in the magma. Thus, in the case of an open vent 

 in which gases were freely given off, hornblende would not form, 

 whereas if the gases were retained for any reason during a period of 

 dormancy, hornblende might develop. 



The close association of hypersthene and hornblende in the rocks of 

 Vanua Levu would indicate that the conditions for the formation of 

 the two minerals are quite similar. It is significant that all the hyper- 

 sthene basalts of Vanua Levu are the results of submarine extrusions. 

 Frank Ferret (1917, p. 45G) has recently described a flow from Strom- 

 boli which entered the sea. He emphasized the rapid formation of a 



