the Secondary Enlargement of Tourmaline. 191 



rocks in which the phenomenon has been noted is not so great 

 that the newly-formed additions, and the grains that do not 

 exhibit secondary enlargements, may not reasonably be attrib- 

 uted to a readjustment of material already existing in the 

 rock as clastic elements, rather than to the introduction of new 

 material. On this hypothesis the frequent and wide-spread 

 occurrence of minute quantities of tourmaline in rocks where 

 its presence cannot be attributed to contact metamorphism, may 

 perhaps be satisfactorily explained. The secondary enlarge- 

 ment of tourmaline seems to have been already observed in 

 similar rocks (see Van Hise, 16th Ann. Rpt. G-eol. Sur., p. 

 685), but 1 have been unable to find any account of it. 



Considering the itacolumites as original sandy sediments 

 derived for the most part from granites and granitic gneisses, 

 other frequent accessories of these latter rocks, and more espe- 

 cially the phosphates, monazite, xenotime and apatite, might 

 be expected to occur, at least sporadically, in them. A careful 

 search failed to reveal unequivocal evidence of their presence, 

 though dubious grains were noted that are suggestive of worn 

 and partially corroded monazite and xenotime. The hypoth- 

 esis that they may have disappeared, giving rise to new com- 

 binations, is favored by the somewhat frequent occurrence of 

 the aluminium phosphate, lazulite, in these rocks, and of 

 amorphous phosphatic masses {favas of the diamond miners) 

 in the detritus derived from them or their associates. To test 

 this hypothesis, a small fragment of itacolumite containing 

 lazulite, evidently authogenic, was examined in the hope that 

 some trace of an original clastic phosphate might still be 

 found. In fact, along with some dubious grains suggestive of 

 altered and corroded xenotime, an undoubted intergrowth of 

 this mineral was detected. There is therefore direct evidence 

 of the association of newly-formed and clastic phosphates in 

 the same rock, which, taken in connection with the strong 

 antecedent probability of the occurrence of allothogenic 

 phosphates in these rocks and their apparent disappearance, 

 and with the fact that monazite and xenotime are frequently 

 partially altered and corroded in the decay of eruptive rocks, 

 gives a reasonable degree of plausibility to the hypothesis. 

 Further observations on this point are however desirable. 



The observations here recorded and others less complete on 

 the argillaceous rocks associated with them, show that with the 

 degree of metamorphism that this series exhibits, all of the 

 original clastic elements except zircon are subject to changes 

 that may obliterate or render dubious their evidence of such 

 origin. In addition to the well-known processes of fragmenta- 

 tion, recrystallization and secondary enlargement by which 

 such evidence may be obliterated, by the renovation of worn 



