CONSIDEEATION OF APPARENT EXCEPTIONS. 137 



to cease.* If, then, these facts will bear the interpretation which I have 

 placed upon them, we have in the rocks themselves the evidence required 

 to show that propylite is a rock which at a certain temperature is just suf- 

 ficiently fused and just sufficiently expanded to fulfill the mechanical con- 

 ditions requisite for eruption. 



It still remains to look at some points in the application of this theory 

 to the succession of eruptions, which would at first sight appear anomalous 

 if not inconsistent with it. 



We do not always find the order of succession heretofore described to 

 have been strictly followed ; we find exceptional cases. Instances are not 

 wanting where true basalts have outflowed prior to the eruption of rhyo- 

 lites, and are even known to be overlaid by trachytes in the Auvergne 

 district of France, or as Lyell has found to be the case in the Madeira 

 Islands. These, however, seem to be exceptional instances. Even in the 

 Auvergne and Madeiras the great preponderance of occurrences conform to 

 the observed law of Richthofen, and so far as our knowledge of other 

 regions extends the departures from this law are not common. But it may 

 be asked whether a single unequivocal exception is not sufficient to seri- 

 ously impair, if not wholly break down, the explanation of the sequence 

 here given. So far are they from impairing it, that I think a. little exam- 

 ination will show that not only ought we to look for exceptions, but we 

 may even be surprised that exceptions have not been found more numer- 

 ous than they appear to be. In the brief explanation given it has been 

 assumed tacitly, that the rise of temperature has been uniform or followed 

 some definite law of variation throughout the entire field of subterranean 

 magmas. In its simplest or typical form the proposition assumes that in 

 all typical or normal cases the rise of temperature aifects all parts of this 

 field alike. But this we could not expect. It is not JDrobable that a uniform 

 rise of temperature would take place in all parts of the field, but may vary 



* It was when I was contemplatiug tho great distauces traversed by sleudcr basalt streams in 

 Southeni Utah that this theory suggested itself to me. I could not doubt that such lavas must have 

 been ejected at a temperature much more than sufficient to melt them. This seemed to contrast pow- 

 erfully with tho habits of trachytic masses. It occurred to me then that this high temperature might 

 be al)8olutely essential to the eruption of so dense a rock as basalt, while a considerably lower one 

 would suffice for lighter rocks. Immediately the higher melting temperature of tho rhyolites and 

 trachytes suggested itself, and almost as quickly as I write it tho theory took form in my mind and tho 

 double function of density and fusibility associated itself with tho double sequence. 



