85 



above that required and which appears in the norm as 

 quartz. By direct experiment it was subsequently shown 

 that most, if not all, of this silica was derived from the 

 agate mortar and pestle which was used in grinding the 

 sample. There is no free silica or quartz shown in the 

 thin sections, nor was any of this mineral found in the 

 separation of the rock by means of the heavy solution. 



The mode or actual mineralogical composition cannot 

 be calculated with certainty on account of the presence 

 of the two micas and the scapolite, the latter having about 

 the same formula as the feldspar. The corundum, mag- 

 netite and calcite are normative, that is, they are present 

 essentially in the percentages given in the norm. From 

 an inspection of the slides the following would seem to be 

 a very close approximation to the mineral composition of 

 the rock: — 



Andesine (near Ab3An 2 ) 72-00 



Nepheline 3 • 00 



Scapolite 2-00 



Corundum (by trial) 13 • 24 



Biotite 5-00 



Muscovite 3 • 00 



Magnetite 1-39 



Calcite -37 



100 • 00 



Owing to the large percentage of corundum present, 

 the rock is a very peculiar and unusual one, and when 

 discovered occupied a new sub-class, order, rang and sub- 

 rang in the quantitative classification. The following 

 names were accordingly proposed for the new rang and 

 sub-rang, and the name dungannonite for the rock itself : — 



Class I Persalane. 



Sub-class II (section I) Dosalane. 



Order 5 Indare [13p.2i7] 



Rang 3 Dungannonase. 



Sub-rang 4 Dungannonose. 



An analysis of the andesine occurring in the rock was 

 made by M. F. Connor. This is given under I. The 

 material for the analysis was obtained by separating the 

 feldspar with Thoulet's solution, but was somewhat im- 

 pure, owing mainly to the admixture of a small amount 



