83 



former in this country, and the Penarfynydd rock as the type of the latter. 

 In both augite and hornblende are present ; but in the former the augite 

 and in the latter the hornblende predominates. Felspar often occurs as an 

 accessory constituent in the picrites, and varies in amount in different 

 specimens taken from the same rock-mass. By an increase in the amount 

 of the felspar and a decrease in the amount of olivine the augite-picrites 

 shade off into the olivine-dolerites, and the hornblende-picrites into rocks which 

 may be conveniently termed oliviiie-diorites. (1) 



In connection with this subject we may refer to a paper by 

 Mr. WILLIAMS (2) on the " Peridotites near Peekskill." The great interest 

 which attaches to these rocks lies in the fact that the mutual relations of the 

 different varieties are well exposed and the rocks themselves remarkably 

 fresh. The allied rocks in the British Isles are known only in boulders and 

 a few more or less isolated exposures, and are usually much altered. The 

 essential constituents of the basic and ultra-basic rocks of Peekskill are 

 brown hornblende, augite (diallage), hypersthene, olivine, biotite and 

 felspar. The accessory constituents are magnetite, magnetic pyrites and 

 pleonaste. The felspathic shade into the non-felspathic rocks by the most 

 insensible gradations, and the different varieties of the latter shade into each 

 other in the same way. The olivine-bearing non-felspathic rocks are divided 

 by Mr. WILLIAMS into two groups ; the one characterised by the predominance 

 of brown hornblende, the other by the predominance of augite or diallage. 

 These two groups shade into each other and into the corresponding felspar- 

 bearing rocks, the olivine-dolerites and the olivine-diorites, by the most 

 insensible gradations. 



Associated with the olivine-bearing non-felspathic rocks are other non- 

 felspathic rocks from which olivine is entirely absent. These consist of two or 

 more of the following minerals: brown hornblende, augite (diallage), hypersthene 

 and biotite. Every possible combination of these minerals appears to be represented 

 in the rocks of this area ; so that the number of varieties produced is practically 

 infinite. 



Here then we find relations similar to those described by Professor JUDD as 

 occurring amongst the peridotites and gabbros of Rum and the Shiant Isles. 

 Nothing can prove more conclusively the impropriety of burdening petrographical 

 literature with names for every slight variety of rock than this evidence that 

 in one and the same rock-mass the number of varieties is practically infinite. 

 The whole rock-mass must be regarded as a mineral- aggregate and 

 described accordingly. 



With reference to the Peekskill rocks it may be remarked that in none 

 of them does olivine appear to rise to anything like the proportion required 

 by TSCHERMAK'S definition of picrite and still less does it rise to the 

 proportion found in the Iherzolites, saxonites and dunites. Thus a variety 



(1) Such for example as some of the diorites described by Mr. ALLPOET, Q.J.G.S., 

 Vol. XXXV., p. 637. It must be remembered that these diorites are associated with rocks 

 which are now known to be Cambrian and not Carboniferous, as was supposed at the time 

 when Mr. ALLPORT'S paper was written. 



(2) A.J.S., 1886, p. 38. 



