Vol. 63.^ CONSTANTS IN MINERALS AND IGNEOUS ROCKS. 153 



(1-517). For better observations of the halo, it was found convenient 

 to use light passed through a coloured medium, such as a solution of 

 potassium-bichromate. 



The following list shows the comparison between the approximate 

 refractive indices of the glasses, with their specific gravities : — 



Xame. Sp. Gr. of Glass. Befractive Index. 



1 . Rhyolite 2*375 Slightly greater than 1 '527 



2. Peterhead granite 2*376 Between 1 '527 and 1-533 



3. Shap granite 2*446 Between 1*527 and 1 -533 



4. Micropoecilitic andesite 2*525 About 1'567 



5. Hyalopilitic andesite 2*550 Between 1*567 and 1*573 



6. Syenite 2*560 „ 1*567 and 1*573 



7. Micropoecilitic andesite 2*570 „ 1*567 and 1*573 



8. Tonalite 2*575 „ 1*567 and 1-573 



9. Penmaenmawr diabase 2*578 ,, 1*567 and 1*573 



10. Rowley-Bag dolerite 2*640 „ 1*573 and 1*578 



11. Guernsey diorite 2*680 „ 1*573 and 1*578 



12. Markfield diorite 2*710 „ 1*573 and 1*578 



13. Olee-Hills dolerite 2*775 Less than 1*578 



14. Gabbro 2*791 Less than 1*578 



15. Whin-Sill dolerite 2*800 About 1 *578 



These results are merely approximate, owing to the wide inter- 

 vals between the fluids ; they comply, however, with the formula 



n 1 



— — =K, where d— Density, n~ Refractive index. 



Fusion of Minerals. 



The methods used in estimating the change of specific gravity 

 which accompanied the fusion of the minerals examined were 

 similar to those employed in dealing with the rocks. Certain 

 members of the soda-lime felspar series were the first chosen to be 

 experimented on, since they are generally recognized as consti- 

 tuting the most interesting and important group of rock-forming 

 minerals. Specimens were selected from known localities, the 

 composition of which had been found by analysis to approximate 

 most nearly to the theoretical composition. 



The method of procedure was as follows : — The mineral to be 

 examined was crushed in a steel mortar, and passed through a 

 fine copper sieve ; the powder thus obtained was placed in a 

 beaker, and washed with distilled water in order to remove the 

 impalpable dust. It was then carefully dried in a water-oven, 

 and finally purified by means of a mineral- separator similar to 

 that described by Prof. Sollas in Quart. Journ. Geol. Soc. vol. lviii 

 (1902) p. 163. The fluids used were carbon-tetrachloride and 

 methylene-iodide. It was found possible, with certain precautions, 

 to separate components differing in specific gravity by two figures 

 in the third place of decimals. 



The chief source of error in this type of apparatus is that due to 

 evaporation round the stoppers. Since the separation will be more 

 perfect the longer the time taken, and the more frequently the 

 fluid is agitated, in dealing with so small a range of specific gra vitr- 

 ei. J. G. S. No. 250. m 



