464 THE CONTINENTAL ORIGIN OF FIJI, 



(Plate xxii., figs. I--). From the bank of the creek there is a long 

 steep slope of talus leading up to the face of a cliff. The lowest 

 exposure in the cliff consists of a well marked bed of conglomerate 

 containing large, perfectl}^ rounded pebbles of a great variety of 

 rocks. Amongst these an undoubted quartz schist occurs, showing 

 that the Fiji area must have undergone considerable dynamic meta- 

 morphism at a period prior to the formation of the conglomerate. 

 Ss far as I am aware, dynamic metaniorphism, powerful enough 

 to give rise to a quartz schist, is confined to continental areas. 

 At all events, we have distinct evidence of a considerable land- 

 mass existing at the time the conglomerate was formed. Overlying 

 the conglomerate is an upraised reef with corals in situ. This 

 has 3^ielded a considerable number of fossils of various kinds, the 

 assemblage of which appears to indicate that the bed is not 

 newer than Pliocene. Conspicuous amongst these fossils are the 

 teeth of a large Carcharodon (Plate xxiii., fig. 3). 



Above the limestone we pass to soapstone, which is here 

 sufficiently coherent to be quarried and used for building stone. 

 At the quarry itself the bedding is just about horizontal, but in 

 the road, about a mile to the north, a southerly dip of about 5° 

 to 7° is encountered. 



This section, like those at Xasoqo and Xadrau (to be described 

 hereafter) gives us positive proof of elevation of the land since 

 early Tertiary time, and also indicates that, before the upheaval 

 took place, areas of very much more ancient rocks were exposed 

 at no great distance. 



From Suva to Xausori nothing but " soapstone " was observed 

 except at Kalabu, where coarse andesite breccia, fairly coarse 

 tuffs, and shell-bearing tuffs occur. 



Ai)out Xausori the alluvial flats are fairly extensive, and 

 alternate with remarkably steep bluffs of " soapstone." The 

 soapstone, which is here at an}^ rate fine redistributed volcanic 

 tuff, is rather carbonaceous. It is interstratified with thin layers 

 of pure river sand from ^ to i inch in thickness. These latter 

 evidently represent the material brought down in flood time. 

 Here and there subfossil twigs are abundant, forming thin layers 



