568 GREAT SKIiPENTINE-BKLT OP NEW SOUTH WALES, V., 



that is difficult to understand, is the almost complete absence of 

 glassy matter among the pyroclastic rocks. As will be seen, 

 they consist (a) of fragments of rocks passed through by the 

 ascending igneous mater-ial, chert, limestone, etc.. and (6) frag- 

 ments, isolated crystals, and portions of crystals from medium- 

 and fine-grained rocks of the dolerite-spilite-keratophyre series. 

 In the case of the "intrusive tuffs," the rock consists largely of 

 grains of minutely crystalline trachytic keratophyre, which recall 

 the constituent granules in the brecciated keratophyre of Hyde's 

 Creek(17). Tliough the ascending magma must have been rapidly 

 chilled by the wet sediments, it must also have been charged 

 with a considerable amount of water, from which it could not 

 free itself. There would thus be a mineraliser constantly present 

 during the period of consolidation, which might partly account 

 for the advanced crystallisation, frequently very minute, of the 

 rock.. Also, any glass that formed in these conditions would pro- 

 bably be rapidl^^devitrified, and, indeed, much of the crypto-crys- 

 talline grains may be devitrified glass. The movement of the 

 molten material below would break up the crust as it consolid- 

 ated, and it would also be shattered by the strains produced in 

 the necessarily rapid variations in temperature, so that, above the 

 level where the crystallisation took place, the comminuted igneous 

 material would move forward in the form of a watery slurry or 

 mud. This mud would escape from the vent in which it rose, by the 

 path of least resistance, which, under a considerable overburden 

 of silt and sea-water, might sometimes be by intrusive injection 

 into the surrounding partially consolidated sediments, at other 

 times, by breaking through and discharging into the sea. This 

 would doubtless be accompanied by more or less energetic con- 

 vulsions. In the latter case, the igneous material would settle 

 down on the sea-floor, as a band of tuff, either pure or mixed 

 with normal marine sediment. Hence might arise the well 

 ascertained fact, that it is often quite impossible, by purely 

 petrographic means, to distinguish between a sedimentary and 

 an intrusive pyroclastic rock in this district. 



An objection to this explanation, which cannot yet be satis- 

 factorily answered, is the uncertainty that the greatest depth 



