DECOMPOSITIOX OF THK ItOCKS. 105 



The atomic ratio deJuciblo fro:n this aualvsis \> IP : 11" : H" -. Si 

 = 0.427 : 1.04 i : 0.8G8 : .3.312. 



l>i;COMPOSITIO\' OF THK CRYSTAI.LIXK TioCKS. 



Character. — All tliu CT v.stalline Tocks are often found in a more or les.s 

 advanced stag-e of decoaiposition. Besides serpentinization, which will Le 

 treated in a separate section of this chapter, there have been other processes 

 at work, particularly tlie epigenetic formation of nralite, cldorite, and nac- 

 rite. Though it is difficult to make out the precise I'elation of tlie.se trans- 

 formations to serpentinization, reasons are given elsewhere for believing 

 them to belong substantially to a different jieriod from tlie serpentinization. 



The conversion of augite to lu'alite is common in the augitic metamor- 

 phics and ordinarily presents no peculiarity. As has already heen men- 

 tioned, liowever, brown hornblende and augite are in one case so a.ssociated 

 as to snggest epigenesis, though it is beheved that the plienomenonisreallv 

 one of envelopment. Chlorite forms directly from brown hornblende, ural- 

 ite, augite, and garnet. It is also not infrequently fonnd in needles in feld- 

 spar in such a way as to suggest the supposition that it may be a result of 

 the attack of feldspar by solutions. Epidote is found much less abundantly 

 than it often is in eruptive rocks. Its relations to chlorite are referred to 

 under the description of epidote. In a single pseudodiabase from New 

 Almaden somewhat irregular, six-sided scales, showing radial striation and 

 remaining sensibly dark between crossed nicols, occur in the feldspars and 

 are supposed to be nacrite. Less well developed flakes of a similar sub- 

 stance are common in other rocks, but no considerable quantity of anything 

 corresponding to the descriptions of kaolin has been detected. Ii'on oxides, 

 carbonates, and leucoxene (probably titanite) are abundant. 



rilTHANITES. 



Character. — Associated with the sandstones of every group in the Coast 

 Ranges is more or less shale^ which, however, seldom forms any large por- 

 tion of the exposures. Some of these shales do not effervesce with acid, 

 and analysis shows that these contain extremely little lime or magnesia ; 

 others are composed to a large extent of carbonates. The shales of the 



