Dr. A. Hubert Cox — Sottth Staffordshire Fire-clays. 61 



The separation of tlie ci-ystalliiie sillimanite, with a specific 

 gravity 3-23, naturally involves a considerable contraction. The 

 result is the formation of a drusy material, in which the cavities 

 are of comparatively large size and quite irregular in shape. 



The marked tendency to the production of sillimanite when 

 substances approximating in composition to kaolin are strongly 

 heated, has been observed on various occasions. Wernadsky ^ 

 obtained it from artificial mixture of AI2 O3 and Si O2, and also from 

 kaolin. He further noted its formation in crucibles and fire-bricks 

 from various fire-clays. It was also observed by Mellor^ in 

 porcelains made from mixtures rich in china-cLiy and felspar fired at 

 temperatures over 1300° C.^ It is a remarkable point that the 

 sillimanite formed from the fire-clays seems not to have the same 

 composition as the naturally-occurring sillimanite of the rocks. 

 The natural product is always stated to have the composition 

 AI2 Si Os, or AI2 O3, Si O2, where the AI2 O3 : Si O2 ratio is as 1:1. 

 The artificial product, however, seems to have the ratio 11:8, as 

 shown bv the following figures : — 



Calculated for 



Si O2 30-07 

 Al, 0„ 69-93 



111 AI2O3, 8 Si02 

 29-96 

 7004 



AI2 O3, Si 0, 

 37-03 

 62-97 



The analysis was carried out by volatilizing the Si O2 from the 

 crystals (obtained as mentioned above, p. 62) as SiF^, the Ala O3 being 

 left as such on ignition. 



A similar result was obtained by "Wernadsky* for sillimanite from 

 fire-clay's, wliereas the sillimanite formed from artificial mixtures of 

 Ala O3 and Si had the normal composition A12 Si Os, whatever the 

 proportions of AI2 O3 and Si O2 in the original mixture.^ 



Sillimanite of the composition 11 AI2 O3, 8 Si O2 was recorded by 

 Sainte-Claire, Deville, and Caron * as formed by the action of Al Fs on 

 Si O2, but, so far as I am aware, there is no case recorded of an 

 analogous product occurring in the metamorpliic rocks. The question 

 is, however, deserving of further investigation. 



Sillimanite is a chemically simple substance, and it is well known 

 that the fusing-poiut of a chemically simple substance is lowered by 

 the introduction of impurities. Now all tiie clays examined contained 

 sandy (quarlzose) material giving an Si O2 : AI2O3 ratio much greater 

 than that in sillimanite. It appeared probable, tliei'efore, that the 

 refractoriness of the clays would be increased by removing the excess 

 of sand. This was actually found to be the case, as already 

 mentioned (p. 60). Since, however, the clays all contain much 



1 Bull. Soc. Min. France, vol. xiii, p. 260, 1890. 

 ■•^ Journ. Soc. Cliem. Ind., vol. xxvi, p. 375, 1907. 



* For a summary of the effects of heat on kaolin and kaolinite, see J. A. 

 Howe, Handbook to the Collection of Kaolin . . . (Mem. Geol. Surv.), 1914, 

 p. 151, with references. 



* Bull. Soc. Min. France, vol. xiii, p. 270, 1890. 

 ^ Op. cit., p. 263. 



" C.R., vol. xlvi, p. 766, 1858. 



