172 HENRY G. SMITH. 



by instituting fresh analyses on material properly purified, or 

 shown by the microscope to be homogeneous. 



They appear to recognise the difficulty in arranging, with any 

 great success, a series of results obtained from analysing impure 

 material ; and the error in this respect has certainly not been 

 rectified by later additions to our knowledge of the composition 

 and formation of kaolin. 



Manifestly the primary basis for a mineral species is that it is 

 in a crystallized or crystalline condition ; and it is by becoming 

 possessed of such, that I have carried out the following investiga- 

 tions, with the hope of being able to assist in the elucidation of 

 this matter. 



In a paper by Mr. J. H. Collins, f.g.s., " On the Nature and 

 Origin of Clays,"* in which he essays to arrange a new formula for 

 Kaolinite; a number of analyses are given, these perhaps are fairly 

 representative, but yet not free from the charge of not being 

 properly purified, as the silica ranges from 41*6 to 49 # 1 per cent. 



Prof. J. D. Dana in his "System of Mineralogy" has given under 

 Kaolinite twenty-six analyses, the results of which are not con- 

 cordant, the silica ranging from 43*13 to 49'1 per cent.; nor is it 

 possible to arrange the formula given for Kaolinite (Al 3 3 2 Si 

 3 + 2 H 2 0) from hardly any of these analyses. 



It is not my intention to discuss the composition of ordinary 

 clay, only so far as it is necessary for the purpose of this paper. 

 The economic value, composition, and quantity, is a matter that 

 can well be left for future discussion. As the clay deposits of 

 New South Wales are extensive, and the clay often of good quality, 

 they will ultimately be in great demand for the purpose of the 

 potter. 



Mr. Collins in his paper (loc. eit.) states that true kaolin, 

 instead of being made up largely of pearly scales as stated by 

 Dana, "is in general, and always when properly prepared, abso- 

 lutely free from them. The hexagonal scales 'which may be 



* Min. Mag., Vol. vn., p. 205. 



