fETROGliAPHY. 331 



These results, though extremely suggestive, are uot, it will be ob- 

 served, in exact accordance with 'those arrived at by Messrs. Biller and 

 Iddings, to whose i)ai)ers we have already referred. The results ob- 

 tained by these gentlemen, coupled with what is already known regard- 

 ing the phenomena of crystallization, seem to show that sti-uctural feat- 

 ures, and to a certnin extent the universal composition of eruptive rocks, 

 are dependent upon (1) the chemical composition of the magma and (2) 

 the varying conditions of heat and pressure under which this magma 

 existed prior to and during the coarse of its eruption. A mineral which 

 separates out under certain temporary conditions of heat and pressure 

 may, under changed conditions, be partially or wholly resorbed, and the 

 condition of chemical equilibrium be so changed as to give rise to crys- 

 talline secretions of quite different nature. 



Mr. G. F. Becker, in a paper treating of the texture of massive rocks,* 

 and drawing his illustrations from the magniliccnt exposure oti'ered by 

 Mount Davidson and the Comstock Lode, Nevada, argues on chemical 

 and theoretical grounds that the structural differences existing between 

 holocrystalline porphyritic and granular locks are due uot to condi- 

 tion of cooling", but rather to original differences in composition and 

 fluidityof the magma, (iranular structure, excepting in rare cases, is due 

 to imperfect fusion and fluidity of the ujagma. Porphyritic structure, 

 on the other hand, is the normal structure of rocks cooling gradually 

 from a high temperature and (;(»ns( (juent state of very ]terfect 13uidity. 

 When the two types of structure arc associated it is argued that this is 

 due to a lack of homogeneity in chemical composition and to tempera- 

 tures sufficient to fuse portions, but not tlie entire mass. Granular 

 rocks as a rule, he argues, have formed at lower temperatures than por- 

 phyries of precisely the samechemical composition. Granular structure is 

 therefore characteristic of rocks formed by the metamorphism of sedi- 

 ments and i)orphyritic structure chara(;teristic of those crystallizing; 

 from homogeneous Huid magmas. These conclusions are based larg'ely 

 on a new law of thernu)-chemistry advanced by Mr. Becker in a previ- 

 ous paper.! In this same connection may be mentioned the researches 

 of Professor J udd, who, in a very interesting'" -and important i)aper in 

 the Geological Magazine for January, 1888,| calls attention, first, to the 

 great dissimilarity in chemical (*omi)osition of rocks classed as enstatite 

 or hypersthene ardesite, the silica i)er(;entages in extreme cases varying 

 nearly 20 per cent, and shows that this difference is due mainly to the 

 relative proportions of the crystalline constituents to the glassy base. 

 Then, treating wholly of the glassy lavas of Krakaloa, he siiows that 

 the obsiilian, (containing a considerable i)ercentage of volatile matter, 

 swells up and fuses at ap])ro.\imately a white heat, while the stony 



* Am. Jonr. of Sci,, 1887, xxxiii, p. 50. 

 t Am. .Joiir. of Sci., 188(), xxxi, p. 120. 



t Tho Natural History of Lavas as illustrated l»y (lui materials ejected froui Kra- 

 katoa. J. \V. Judd. Gcol. May,., .January, 1~H,-S, vol. v, ]>. 1. 



