584 GREAT SERPENTINE-BI£LT OF NEW SOUTH WALES, V., 



]^roximity of land, (3) tlie absence of any coarse terriginous sedi- 

 ment, and (4) the intercalation of coral-limestones. The present 

 writer concurs entirely with these conclusions. The only considera- 

 tion which gave rise to some doubt was the possibility that the 

 Baldwin Agglomerates, now proved to be interstratified in the 

 radiolarian series, might owe the rounded character of their inclu- 

 sions to water-erosion. The present study, however, tends to the 

 conclusion, that, even if this were so, as seems possible in some 

 cases, this would not indicate the presence of a persistent coast- 

 line, but only the development and rapid destruction of islands, 

 which were the summits of masses of agglomerate erupted from 

 vents in the flat floor of the sea. The development of the Baldwin 

 Agglomerates marks an epoch when such great eruptions were in 

 progress. It is not yet clear that the building of volcanic islands 

 characterised the eruptions, that produced the Igneous Zone in tlie 

 Middle Devonian Series. The eruptions may have occurred at a 

 considerable distance from a persistent coastline. Except for the 

 products of these great eruptions and the many minor convulsions, 

 the sediments are of the finest grainsize, and largely composed of 

 the remains of radiolaria. The depth of the sea in which they were 

 laid down, must have been sufficiently great to give the overburden 

 requisite for the production of intrusive tuffs, and sufiieiently 

 shallow to permit of the formation of ripple-markings, and coral- 

 reefs. The exact depth at which this balance was obtained cannot 

 be estimated with any precision. Doubtless it varied somewhat, as 

 the presence of definite zones of limestone would lead one to infer ; 

 but as the radiolarian rocks, with interbedded and intrusive tuffs, 

 adjacent to the limestones, are quite similar to those elsewhere in the 

 series, there is no reason to assume that these variations were large. 

 We may, perhaps, conclude that there is no evidence that the radio- 

 larian rocks were formed at a depth less than the maximum at 

 which it was possible for the limestone to liave been formed. The 

 work of Darwin, Dana and many others has shown that the reef- 

 building corals in the modern seas, do not live at greater depths 

 than about fifteen or twenty fathoms, though certain isolated forms 

 extend much further down. The corals of the past seem to have 



