Page images

notice of the other four garnet molecules' detected by me, which, though usually absent or rare, are sometimes present in notable proportions.

Boeke's diagram of variation in the composition of garnets has been arrived at by plotting analyses. It expresses, therefore, a statistical fact, namely that the majority of garnets can be grouped into two sets or series. The question is whether this is due to the

[merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][ocr errors][merged small][merged small][merged small][merged small]
[ocr errors]



[blocks in formation]

Fig. 1. Variations in composition of garnet after Boeke, Zeit, Kryst., LIII, p. 149.



inherent physical properties of garnets rendering impossible the existence of mixtures' occupying the field between 'pyralspite ' and 'ugrandite', or is merely due to the opportunities of association of the various elements usually provided by Nature. Boeke's diagram does in fact show several garnets occupying this intermediate terrain; but the author discusses each of these exceptions in turn and gives reasons (impurities, imperfect analysis, etc.), for rejecting them.

I have, however, myself published analyses of several garnets that fall well into the terrain between 'pyralspite' and 'ugrandite '1. Referring to the table on page 341 of the paper cited, the garnets from Chargaon (magnesia-blythite), Sakrasanhalli (calcspessartite), Kotakarra (calc-spessartite), Garbham (spandite), and Kodur (spandite), all fall well into this terrain, and there seems to be little doubt that complete solubility exist between some portions at least of the 'pyralspite' group and some portions at least of the 'ugrandite' group, and particularly between spessartite and andradite; and that there is a manganese-lime series of garnets cutting diagonally across Boeke's diagram from north-west to southeast. Bocke's diagram is reproduced here with the addition of six spots corresponding in numbers to six analyses in my own paper, of five spots corresponding to five analyses in Fleischer's paper, but distinguished by the letter F. In addition there is a high degree of mutual solubility between almandite and grossularite, as instanced by analyses Nos. 13 (gralmandite) and 25, 20, 15, 21, 17 and 18 (magnesia-gralmandite) of M. Fleischer's recent paper2, and by the analyses by N. Jayaram of Nellore garnets3. The six Nellore garnets contain 8.55 to 18.52 per cent. of grossularite, 7-50 to 21.10 per cent. of pyrope, and 56-72 to 64-72 per cent. of almandite; and as a group they may be described as calc-pyralmandites and magnesia-gralmandites according to the relative proportions of pyrope and grossularite. Analysis No. 13 of Fleischer's paper reveals the presence of 53-56 per cent. of garnets of the 'pyralspite series and 46-43 per cent. of the ugrandite' series. It seems certain in fact that besides the 'pyralspite' series and the ugrandite' series that there are a

[ocr errors]

1' Note on the Manganese-Lime Series of Garnets', Rec. Geol. Surv. Ind., LXVIII, pp. 337-343, (1934).

2 The relation between chemical composition and physical properties in the garnet group, Amer. Min., Vol. 22, p. 754, (1937).

The mineralogy and chemical composition of garnets from the schist-complex of Nellore Proc. Ind. Acad. Sci., V, Sec. A, pp. 148-160, (1937).

spandite series and a gralmandite series cutting across the space between the two former series. Further, amongst the garnet analyses listed by Fleischer there is not one which represents a garnet to which either of the terms pyralspite or ugrandite could be applied on a basis that required the presence of a minimum of 10 per cent. of a constituent for recognition in the name.

Other garnets that fall outside the fields of 'pyralspite' and ugrandite' are the three ferric garnets khoharite, skiagite, and calderite, which occupy the north-east corner of the diagram, the position also for blythite if manganic garnets be grouped with the ferric garnets.

Further, the atomic structure of garnet does not appear to supply any structural reason for the supposed limited solubilities between the two chief groups statistically detected by Boeke. All the garnets belong to the space-group 010 or Im3m which is the space-group of the highest degree of symmetry amongst the 230 possible groups. It is true that the physical properties of garnets-refractive index and specific gravity for instance-vary with their composition, as has been well shown by W. E. Ford1 and M. Fleischer2, but these variations are merely the result of variation in the proportions of the various divalent and trivalent atoms in the space lattice. There seems to be no reason based on the size of these atoms why any particular association or admixture' should be physically incompatible with the garnet structure.

And as we have the fact of two series of garnets, the manganeselime garnets (spandite series), and the lime-iron garnets (gralmandite series), cutting right across the space between 'pyralspite' and 'ugrandite' in Boeke's diagram, it seems more logical to assume that the infrequency of garnets the analyses of which occupy this. space is a reflection of the limited opportunities provided by Nature based on the proportions of the elements available for forming garnets, rather than of any inherent difficulty based on the atomic structure of garnet and the physical properties dependent thereon.

In order to illustrate the suitability or otherwise of using hybrid and hyphenated names for garnets we may apply the methods. noticed in this paper to the batch of analyses of garnets, 57 in number, collected in Fleischer's paper to which reference has been made above, adopting as a rough limit that a minimum of 10 per cent,

1 Amer. Journ. Sci., Vol. 40, pp. 33-49, (1915).

2 Amer. Min., Vol. 22, pp. 751-759, (1937).

[ocr errors]

of a garnet should be present to warrant recognition in the name, and that of those above this limit the two major garnets should be conjoined in the name, and the third (if important enough) should be recognised by a chemical prefix:

[merged small][merged small][merged small][merged small][merged small][ocr errors][ocr errors][ocr errors][ocr errors][ocr errors][ocr errors][ocr errors][ocr errors][ocr errors][merged small][merged small][merged small][merged small]

These names do not, of course, possess any quantitative significance. Thus the garnets included above under spalmandite range from No. 51, with 75.00 per cent. of almandite and 15-63 per cent. of spessarite, to No. 40, with 29-24 per cent. of almandite and 69-35 per cent. of spessartite; whilst under grandite we have No. 3, with 83.96 per cent. of grossularite and 10-80 per cent. of andradite, and No. 54, with 17.05 per cent. of grossularite and 73-70 per cent. of andradite.

In this discussion on the nomenclature of garnets I have not thought it necessary to discuss schorlomite, which is treated by Dana as a separate mineral from garnet with the formula 3Ca0. (Fe,Ti) 203-3(Si, Ti)02, from which, combined with the fact that garnet is isometric, it seems that we may regard schorlomite as a titaniferous andradite, which has this peculiarity, judging from the analyses of specimens from Magnet Cove given by Dana, that it is freer from the presence of other garnet molecules' than most garnets.

It has also seemed unnecessary to discuss varietal names such as polyadelphite, rhodolite, and topazolite, applied to garnets from particular localities.

[ocr errors]


"The Mineral Production of India" is compiled annually in August for the previous year and is published, usually in October, in Part 3 of the Records of the Geological Survey of India.

Owing to the very varied sources from which these statistics. are derived and the delays, not attributable to the Geological Survey of India, in obtaining them, it has been found impossible to produce this detailed review earlier in the year. It has, however, been felt that a less complete and less accurate but substantially earlier issue of statistics relating to a few of the more important minerals produced in India would be of interest to the public.

These figures must be considered as provisional and partial only; the revised and complete statistics will be issued as usual in Part 3 of this volume. The figures in brackets give the final productions for 1936.

Antimonial lead




[ocr errors]

The production by the Burma Corporation Ltd., amounted to 1,150 tons in 1937. (1,240 tons). This product contains 81.66 per cent. of lead, 17.69 per cent. of antimony, 0.21 per cent. of copper and 3.44 ozs. of silver to the ton.

The production in Bihar amounted to 6,197 tons, (7,053 tons), in Baluchistan 27,209 tons, (21,089 tons), and in Mysore State 26,013 tons, (21,344 tons).

22,313,205 tons of coal were raised from the coal mines worked under the Indian Mines Act. (20,585,691 tons).

See Nickel Speiss.


« PreviousContinue »