THE EVOLUTION OF PETROLOCIIOAL IDEAS. 295 



nionre to form, and this order may or may not a^roc with that in which 

 they cease to form. The lawsyvhich expi-ess the ordei- of formation 

 of minerals and the cheniical and physical conditions which conti'ol 

 that ordor ha\'e not as yet Ikmmi detinitcdy established. 



One of the most important papers on theoretical peti'ology is un- 

 doubtedly that by Professor Rosenbusch on the signiticance of the 

 granular and porphyritic structures in massive rocks. The importance 

 of this paper must not be judged simply by the amount of truth in the 

 principles enunciated, but rather b}^ the stimulus which it gave to 

 theoretical considerations and to researches directed toward a partic- 

 ular end. The constitu(Mits of massive I'ocks are di\'ided by Professor 

 Kosenbusch into four groups: 



1. The ores and accessory constituents (matiiictitc, liamatitc, ilmenitc, apa- 



tite, zircon, s])inel, and titanite). 



2. The I'erroinagncsiaii constituents (l)iotiti', lnn-nlilt'iidf, i>yi-()xcnt', aiid 



olivine). 

 ?>. The felspathic constitticnts ( rt'lsi»ar, uephelin*'. Incite, TncliUte, sndalite, 



and iiaiiyne). 

 4. Free silica. 



Professor Kosenbusch pointed out that members of the iirst group 

 precede those of the other grou])s; that in granites and syenites the 

 members of the second group precede those of the third; but that in 

 the diabases and ga))))r()s the order is inverted, and that in l)oth groups 

 silica is the last. The general conclusion is reached that ""the order 

 of consolidation of the silicates and, consequently, their crv>!tallo- 

 graphic development (idiomorphism), corresponds to a law of decreas- 

 ing basicity; the ores and accessor v minerals are the earliest, and 

 quartz is the latest, product of the rock-forming process,'" 



This empirical law expresses, in a broad and general way, the main 

 facts observed with regard to the sequence of minerals in the large 

 and important group of intermediate rocks, but it breaks down when 

 applied to the most acid and the most basic rocks; quartz is often 

 formed before felspar in the former, and iron ores are not inf re((uently 

 formed after felspar in the latter. 



The views that we hold regarding the laws which express the order 

 of consolidation in igneous magmas will necessarily Ije colored by our 

 conceptions as to the nature of these magmas. A great advance in 

 the evolution of ideas on this subject is marked by a short letter, 

 wi'itten by Bunsen to Streng, and published in tlu^ Journal of the 

 (Terman Geological Society for ISOI. In this letter Punsen })oints out 

 that the arguments against the igneous origin of granite, so far as they 

 rest ujK)n the so-called anomalous order of consolidation of the miner- 

 als, are based on a misconception of the nature of the process of con- 

 sohdation. He says, 



"The temperature at which a substance consolidates from a state of 

 fusion is never that at which it separates from a solution in aiKjther 



