Page images

Contributions to the study of the Pyroxenic varieties of Gneiss and of the Scapolite-bearing Rocks1; by M. Al. Lacroix. Ceylon and Salem. (Translated by F. R. MALLET, late Superintendent, Geological Survey of India)

Translator's Re?4arks.

The memoir, with the above title, appeared in the Bulletin de la SociM Franfaise de Mineralogie for April 1889.' As justly remarked by the author, the portion translated in the following pages gives the result of the first detailed microscopical work that has been devoted to the crystalline rocks of Ceylon and Salem. The reason for this is not far to seek. Of the different writers, quoted by M. Lacroix, who have occupied themselves with the geology of portions of Southern India, the earliest wrote in 1802, the latest in 1864—all of them, therefore, at a time when the modern science of micro-petrology was non-existent.

M. Lacroix's memoir being, in the main, devoted to certain special types of rock only, his work relating to the Indian region necessarily leaves a large field still untouched. I understand that it is his intention to enter upon this in a second paper at some future period. In the meantime, it is a distinct gain for Indian geology that the rocks within the scope assigned should have been studied in such detail by a petrologist so eminent as the author of 'Les Miniraux da Roches' It will be seen that the results obtained are both interesting and important, the occurrence of several remarkable types of rock having been established, and numerous additions made to our knowledge of Indian mineralogy. In how far the chronological classification suggested by M. Lacroix, founded on petrological analogies with the crystalline rocks of Western Europe, will bear the test of extended observation in the field, is a point that must be left to the geologists of Madras. That the problem is very far from a simple one appears from the observations hitherto made by the Geological Survey.

With reference to the following translation, there are one or two points that it may be well to allude to here.

While it was in hand, M. Lacroix furnished me with several additions, and a few alterations, which are distinguished from the original text by insertion in square brackets [ ].

While French mineralogists generally apply the term 'weme'rite' to the group of closely related minerals, and 'scapolite' to the most prominent species, English authors, although not all in accordance with each other, most commonly use the names in the reverse sense. In the following pages, therefore, wernerite is translated 'scapolite,' and vice versd.

The crystallcgraphic formulas are given in the original according to both LeVy's system and Miller's. As the former of these is, as a rule, less familiar to English mineralogists than Naumann's, the latter has been adopted in the sequel.

1 Contributions a I'e'tude des gneiss a pyroxene et des roches a wernerite.
* Vol.XII, No. 4, p. 83.

The three principal indices of refraction, and aiso the three axes of optical elasticity of crystals, are represented in M. Lacroix's memoir by the letters ng, nm, np, which are translated as follows :—

ng = 7 or = c
rim = /3 or = h
np = a or = a

The latter part of the' Summary and Conclusions,' on " The geological distribution of the Scapolites," has not been translated, as it is entirely of a general character, without special reference in any way to Ceylon or India.

In conclusion, I must add that the thanks of the Geological Survey are due to M. Lacroix for kindly placing the original zinc-plates at its disposal. Hence the illustrations in the following pages are identical with those in the memoir as first published. TABLE OF CONTENTS.

% Paoi


Chapter I. — Pyroxenic and scapolitic gneiss of Brittany.1
Chapter II.—Central Europe.1

Chapter III.—Spain, Algeria, West and East Africa.1
Chapter IV.—Scapolitic rocks of Norway.1
Chapter V.—Scapolite-bearing pyroxenic gneiss of Sweden.1
Chapter VI.—United States.1

Chapter VII.—Gneiss of Ceylon and of the District of Salem ..... 158
Historical; geographical distribution . 158

Stratigraphy . . . . . . . . , -159

Acid gneisses —

(a) Biotite and Sillimanite-gneiss ...... 161

Excptional rocks . . . . . . . .162

(A) Qarnetiferous leptynites ....... 166

Pyroxenic leptynites ....... ,67

(c) Granulitic microcline-gneiss ..... 168

Pegmatites . . . . . . . . . .170

Basic gneisses—

(a) Pyroxenic and hornhlendic gneiss ..... 173

(6) Pyroxenic and hornhlendic gneiss . . . . • 17S

Anorthite-gneiss . . . . . . . .183

Wollastonite and scapolite-bearing rock . . . .189

Cipolins . • . • • ■ . . . .191

Mica-schists .......... !o/>

Pyroxenic gneiss of New Caledonia.1
Scapolite-bearing pyroxenic gneiss of Brazil.1

Chapter VIII.— Summary and conclusions— ,98

Geological distribution of the Scapolites.1

Explanation of the Plate ....... . 2oo


In the upper part of the gneissose series (Ft/age des gneiss), in numerous regions, pyroxenic rocks occur which present a very peculiar facies. These rocks, often associated with hornblendic gneisses and amphibolites (amphiboUies), have a granular structure more or less fine-grained; their mineralogical composition is generally very simple, but admitting of a number of varieties, due to the frequent variations in the chemical composition, and relative proportions, of the lightcoloured constituent minerals (elements blancs). These rocks, by some authors designated pyroxenites, are in this memoir called pyroxenic gneiss (gneiss a pyroxene). We will reserve the term 'pyroxenite' for rocks composed exclusively of pyroxene.

One of the characteristic traits of this class of gneiss is the frequent presence of minerals of the scapolite group. They play the same part as the felspars, which they accompany or replace. My researches show that the scapolites are very abundant in rocks in which they have been for a long time overlooked. I have been able to add a certain number of new occurrences of such rocks to those formerly known.

1 Not translated.


No comprehensive memoir having been yet published on the scapolite-bearing rocks, I have added to the present one a description of two dipyre-bearing rocks coming, one from Norway, the other from Algeria. The dipyre-diorite of Norway being the result of the transformation of an olivine-gabbro, I have been led to study this latter rock, and, in consequence, a French occurrence of olivine-gabbro that I have discovered in Loire InfeVieure.

Most of the rocks described have been collected by myself, during the course of work which has been entrusted to me, either by the Minister of Public Instruction, or by the ' Exole des Hautes Etudes.' Generally speaking, I have devoted myself to the study, not only of the pyroxenic gneisses, but also to that of the rocks which are intimately associated with them in the localities visited.

This work is only the prelude to a comprehensive study of the basic gneisses that I propose to carry out. I must here express my best thanks to MM. des Cloizeaux, Fouque' and Michel-LeVy, who have aided me with their advice, and to the Sociiti MMralogique, which has undertaken the publication of this memoir.

Chapter VII.—Gneiss Of Ceylon And Of The Salem District.

Historical; and Geographical distribution.

The rocks which have served as the basis for the present chapter are preserved amongst the collections of the 'College de France' and of the 'Museum d'Histoire naturelle'1; a portion of them was collected in 1819 by Leschenault de la Tour, who travelled on behalf of the Museum. Some of them also formed a part of de Bournon's collection, preserved in the 'College de France,' and are the originals which that savant made use of for his monograph on Corundum.8

The Ceylon rocks came from the region which extends from the town of Colombo, on the west coast of the island, as far as Kandy.

The rocks of the Indian peninsula were collected in the district of Salem (Madras Presidency).

The town of Salem is built, at an altitude of 900 English feet, upon a plateau watered by the Cauvery, and dominated to the north by the Shevaroy mountains, which rise to [an average elevation of 4,500 feet, the highest point being 5,4io].

The rocks and minerals included in de Bournon's collection are labelled as coming from Salem "on the Coromandel Coast," although Salem is over 100 miles from the sea, to the W. S. W. of Pondicherry.

Occupied, as he mainly was, with botanical and zoological pursuits, Leschenault has given very few stratigraphical details concerning the region he travelled through.'

1 Thanks to the kindness of MM. Dsubree and St. Meunier, I have had full opportunity for studying the specimens in the Museum collections.

* " Description of the Corundum stone and its varieties."—Philosophical Transactions, London, (1802), p. 232.

• [" Relation d'un voyage a Karikal ct h Salem."— Mem. du Museum d'hist. natur. de Paris VI, (1820) 329-348.]

The rocks from the neighbourhood of Salem, and those from Ceylon, present a striking analogy from a petrographical point of view. As far as we know, they have never been the subject of any petrographical work.

De Bournon 1 confines himself to a description of the minerals accompanying the corundum. Newbold,2 Campbell3 and Balfour*, devoting their attention to the same subject, are still more cursory.

In 1864 William King and Robert Bruce Foote published a memoir "On the geological structure of portions of the districts of Trichinopoly, Salem and South Arcot, Madras "6 which gives some geological and stratigraphical information respecting the region near Salem. The gneissic rocks in question, however, are not treated on very fully.

In Greenough's maps, and in that more recently published by the Geological Survey of India,6 the region with which we are concerned is tinted uniformly as metamorphic.

The 'Manual of the Geology of India'7 does not enter into more detail, giving merely references tos the preceding memoirs.

To the south of Salem the Cauvery is bordered by small sterile eminences without trees. The Salem plateau is formed of small undulations covered with jungle until the Baramahal plateau is reached, and the Shevaroy mountains which rise above it.


King and Foote describe, in the Salem region, several varieties of quartzose or hornblendic gneiss, the latter of which are associated with cipoiins; these rocks are traversed by dykes of porphyrite.

Hornblendic rocks predominate, and one of them, designated " Syenitoid gneiss," is the oldest rock in that part of the country. Limestone is of rare occurrence.

The order of succession of the different rocks enumerated is not given.

Judging by means of analogy with the gneissic rocks of the regions that we have already described,9 we shall indicate, broadly, the probable order of succession of these diverse rocks, or, at least, we shall point out the existence, in the area under discussion, of rocks that should be classed with f1, and x of the detailed geological map of France.

From a petrographical point of view we may distinguish two series, that of the acid, and that of the basic, gneisses.10

'op. at.

J]ourn. Roy. As. Soc, VII (1842), 219.
1Cal. Journ. Nat. Hist., II (1842), 281.
•Srlec. Rec. Govt. Madras, XXXIX (1857), 91.
'Memoirs of the Geological Survey of India, I J (1864).
•Preliminary sketch of the Geology of India, Calcutta (1877).
1I and II, Medlicott and Blanford; III, Ball; IV, Mallet

"It would be more correct to say "condensed accounts of the most important information contained in."—F. R. M.

•In the chapters not translated.—F. R. M

10 We do not include the serpentinous and dolomitic rocks forming the chalk hills near Salem, that the preceding authors have spoken of, but of which we have no specimens at our disposal.

« PreviousContinue »