PROCEEDINGS OF SECTION C. 



191 



I have shown that, on determining the position of the analyses of 

 Australites in the Quantitative Classification it was clear that they 

 exhibited a decided relationship in chemical composition. This re- 

 lationship is upheld by the classification of the two fresh analyses. 

 The analysis of the specimen from Uralla falls into the sub-rang Almerose, 

 while that from Curdie's Inlet falls into an unnamed sub-rang for 

 which I have suggested the name of Curdieose. 



It has been pointed out that all the analyses of Australites, Billi- 

 tonites, and Moldavites fall into groups poorly represented among 

 terrestrial rocks, and consequently they are considered to be abnormal 

 in composition. 



Mr. E. J, Dunn disagrees with this conclusion and writes* " the 

 resemblance of Australite to Obsidian is so great as to practically 

 establish the identity of Australites and Obsidian in composition, and 

 therefore any argument founded on their differences of composition 

 fails." 



Mr. Dunn bases his arguments on the ranges of variation of the 

 percentage of the different oxides. This method of comparison of 

 analyses tends to obscure the truth rather than to throw aujjlight on the 

 matter, and arguments based on such ranges of variation may be 

 regarded as practically valueless. 



Another method of comparing the analyses is by means of a 

 Variation Diagram. This is constructed with silica percentage as 

 abscissae, and the percentages of the remaining oxide as ordinates. 



As the analyses have been done under various conditions they are 

 not in the most suitable form for comparison. Thus titanium dioxide 

 and manganous oxide were determined in some cases and not in others, 

 and in one case the ferrous iron was not determined. To obviate 

 errors as much as possible the analyses have been recalculated to 100 

 per cent., the water being omitted, the total iron calculated as ferrous 

 oxide, the titanium dioxide added to the alumina, and the manganous 



• Op. eit., p. 10. 



