1893. J T. H. Holland — Petrology of Job Char-nock's Tombstone. 163 



ed, although the mineral has been frequently found as a constituent of 

 the intermediate, basic and ultra-basic noncrystalline rocks. The 

 precise reasons "why the micas, hornblendes, and, more rarely, augites 

 should occur as the ferro-magnesian constituents of granites, and not 

 hypersthene, have never been accurately settled. The discovery of 

 hypersthene, therefore, in this capacity fills a very well-marked gap 

 in the granitic series, and for the time we can do no more than record 

 as precisely as possible its nature and mode of occurrence, with the 

 hope that in future the facts may be of service in framing an hypo- 

 thesis for explaining the fact that chemically similar magmas, under 

 special conditions of temperature and pressure during the process of 

 consolidation, give rise to different mineral species. 



(5). GARNET of the almandine variety occurs very sparingly in 

 the rock, and seldom shows anything approaching idiomorphic crystal- 

 line form. 



(6). MA G\N"ETITE in small grains is sparsely scattered amongst 

 the other minerals. 



The rock has a specific gravity of 2'646, agreeing thus with normal 

 granites. 



In microscopic and macroscopic characters this rock agrees with 

 certain specimens which I have recently collected in the Madras Pre- 

 sidency. At different places in the south of India (Pallavaram in the 

 Chin gleput district, the Shevaroy and Nilgiri hills, in N'.-W. Madura, 

 and in Travancore) there occur exposures of igneous rock in which 

 hypersthene is a constant constituent, and which at the same time 

 exhibit every gradation in acidity, from hypersthene-granite, the most 

 silicious (acid), to pyroxenite the most basic. These rocks, although 

 their exposures are now separated by such distances from one another, 

 I believe to have been derived from a common molten magma : they 

 belong to one " petrographical province," and the differentiation of the 

 originally homogeneous molten material into masses so widely distinct 

 in chemical composition can be shown to be in agreement with well-es- 

 tablished, though recent, physical principles. 



The massive rocks of the Nilgiri Hills, and the Shevaroys, as well 

 as the similar rocks found in the localities mentioned above, have been 

 hitherto regarded as belonging to the great metamorphic series of the 

 South. Observations made during recent visits to the Madras Presi- 

 dency have, however, convinced me that this series, together with 

 certain others not now under discussion, must be looked upon as in- 

 trusive igneous rocks of younger age than the normal gneiss. 



The evidences for these conclusions I hope shortly to produce in 

 detail. For the present, however, we are concerned in identifying 



