1 1 6 GEOLOGY OF THE HIGH PLATEAUS. 



sistent with the assumption that lavas are portions of a primordial, uncon- 

 gealed earth-liquid, forming either a general fluid nucleus or extensive iso- 

 lated vesicles. They point rather to many small reservoirs, situated at no 

 very great depths, each of which contains, not a primordial liquid, but a 

 liquid secreted, so to speak, from surrounding rocks, or generated by a sec- 

 ondary and progressive fusion of solidified matter occurring in macula within 

 the layers of the rocky envelope of the earth. The whole tenor of volcanic 

 phenomena bespeaks a process which is extremely local — a process which 

 has an inception, a growth, a culmination, a decadence, and a final cessa- 

 tion, all within a limited and rather small area and determined by some 

 local cause. 



But we find the strongest evidence against the hypothesis that lavas 

 are primordial liquids when we come to the study of their physical, chemi- 

 cal, and mineralogical characters. We do not, indeed, have any very deci- 

 sive grounds for asserting what the primordial liquids might consist of or 

 what would be their petrographic characters if any of them were erupted 

 to the surface, and so far we might not be justified in saying that the lavas' 

 from volcanoes are distinct from them. But there are some eruptive masses 

 which are verj* plainly not primordial. For instance, a decidedlv conspicu- 

 ous mass of these products are not fused rocks, but hot mud holding large 

 quantities of rocky fragments, which have unmistakably formed the clastic 

 components of strata. The volcanoes of Central America and the Andes 

 and of the Batavian Islands have within the last century disgorged astound- 

 ing masses of hot mud — material which has not been fused at all, but 

 rendered plastic and capable of flow by the combined action of heat and 

 watery solution. It cannot be admitted that such erupta can have come 

 from primordial materials. And the indications are no less distinct that the 

 greater part of the true lavas have originated from other sources. 



The careful and systematic study of the petrographic characters of all 

 rocks, whether sedimentary, metamorphic, or eruptive, has enabled us to 

 compare them intelligently, and to form some conclusions as to the homolo- 

 gies on the one hand and the distinctions on the other which exist between 

 them. The great generalization that the foliated crystalline rocks are 

 altered sediments has long since passed into geological science as a fully 



