TF. Camphell Smith — Borolaniie from Ross-shire. 155 



Three sections were cut from the remains of the specimen analyzed, 

 and these were examined with a view to getting some idea of the 

 proportions of minerals present. 



The garnets are qnite typical, pale brown, and filled with inclusions 

 of small sphenes, some biotite and a very little apatite. The larger 

 crystals (2mm. diam.) are very irregular and broken; the smaller 

 (•1 mm. diam.) show excellent sections of rhombic dodecahedra. 

 !No sphene occurs outside the garnets ; apatite is sparingly present, 

 and green pleochroic segirine was observed only as two small patches 

 in one section. The biotite is dark green, occurring as somewhat 

 I'agged flakes in the orthoclase-melanite aggregate and to a small 

 extent as inclusions in the garnet. 



' Dactylotype ' structure was well shown, the material forming it 

 being apparently of lower refractive index than orthoclase. Here and 

 there rather indefinite black crosses, shown by parts of the zeolite 

 between crossed nicols, were observed. 



Estimates of the volume-percentages of the minerals of each of the 

 three slices were made by three independent observers, making nine in 

 all. The mean of the nine is : Garnet, 25-3 ; orthoclase, 51 ; ' Spreu- 

 stein,' 19; and biotite, 5. Taking the specific gravities of these 

 minerals as 3-6, 2-57, 2-3, and 2-9 respectively, the percentage mineral 

 composition becomes: Garnet (including sphene), 32-48; orthoclase, 

 46-75; ' Spreustein,' 15-59; and biotite, 5-17. 



The specific gravity calculated from these values, assuming the 

 minerals to have the specific gravity above given, is 2-79. The actual 

 specific gravity was found to be 2" 77. 



In view of the fact that the only minerals present, whose com- 

 position is known, are orthoclase and sphene, and further that there 

 is very little guide as to the composition of the zeolite, any definite 

 calculation of mineral composition from the analysis is extremely 

 difficult, and it is probable that several equally plausible arrangements 

 might be made. The figures which seem to fit best are here given : — 



A fair degree of approximation is observed between the calculated 

 and estimated results. 



