156 TT. Camphell SmitJi — Bo)'olan'itc from JRo8S-sIiire. 



The melanite is assumed to contain all the Fe^ Og, and to have the 

 composition of a melanite from the nepheline-syenite of Duugannon, 

 Ontario (after depriving it of sphene and water). A thomsonite is 

 selected because it is the only zeolite which will bring about the 

 proper ratio between ^'ajO-j-CaO, and water. In view, however, 

 of the rather uncertain condition of the water in zeolites, this cannot 

 be regarded as a strong piece of evidence that the zeolite has this 

 composition. The biotite is purely hypothetical, and doubtless should 

 contain some water. 



The dyke-rock of Camas Eilean Ghlais seems to have very little in 

 common with borolanite. A rough calculation of the 'mode' shows 

 it to contain about 20 per cent, of nepheline, 15'5 per cent, melanite, 

 42 per cent, orthoclase, with some oegirine, nosean, and biotite. 

 A glance at the two analyses given below shows the great difference 

 betvfeen the two rocks — 



It seems very difficult to find any rock whose cumposition is very 

 near that of borolanite. 



A calculation of its true position in the quantitative classification 

 of Cross, Iddings, Washington, and Pirsson shows that it falls in 

 Subrang 3, Eang 2 (Vulturase), Order 7 of Class II. No other 

 analyses seem to fall in this subrang. The dyke-rock falls in a similar 

 position in Order 6 of the same class, and therefore is in a sense 

 nearly allied to borolanite. The ' norm ' of borolanite shows very 

 little similarity to the ' mode ' or true mineral composition, orthoclase 

 being the only mineral which occurs in both. The ' norm ' is as 

 follows : orthoclase, 30-86 per cent. ; anorthite, iri2 ; leucite, 

 13-32; nepheline, 15-62; woUastonite, 16-70. 



Taking the molecular ratios of Si O3 and alkalies as abscissoe and 

 ordinates respectively, about a hundred analyses, mainly of plutonic 

 rocks, were plotted on squared paper. The points fall roughly into 

 the two well-known curves ' alkaline ' and ' calc- alkaline ', and it was 

 hoped that the rocks closely allied to borolanite might be thus found. 

 The result is somewhat disappointing. Borolanite falls on the upper 

 curve, and occupies a position intermediate between the nepheline and 



