132 ANNALS NEW YORK ACADEMY OF SCIENCES 



established mineral species is fixed, the possible object of analysis of a 

 specimen identified by other means may be two-fold : determination of any 

 replacements of components in the chief mineral ; and demonstration of the 

 constitution of other minerals which may be intermixed in the aggregate. 

 The latter may be of great importance in elucidation of genetic history and 

 relationships of the chief mineral. 



Several methods have already been devised and applied toward quanti- 

 tative determination of the elements of such intergrowths or aggregates: 

 such as the graphic methods for measurement of their respective areas in a 

 microscopic field, by means of drawing or photography; that of separation 

 of the elements in a crushed aggregate by suspension in a dense liquid; 

 that of separation of ferruginous minerals from a pulverized aggregate by 

 means of an electromagnet; that of separate chemical analysis of the por- 

 tions of an aggregate soluble and insoluble in an acid; and that of com- 

 parison of the simplified bulk analysis with a series of hypothetical chemical 

 compounds. The first two methods are inapplicable to aggregates whose 

 granulation is microscopic; the next two are limited and imperfect, through 

 dependence upon a single character, and the last is subject to the errors 

 usual to excessive reliance upon hypothesis rather than upon data of obser- 

 vation. 



A more simple and effective method, in many cases, is that shown in the 

 practice of the early mineralogists. Within every chemical analysis of a 

 mineral substance lies the Key to its constitution. For its completion a 

 re-arrangement or recasting is needed to determine the existing minerals as 

 combinations of stated components. This can be carried out where the 

 data are fairly complete, sometimes with great ease, and the results tend 

 toward solution of long mooted problems and elucidation of the character 

 of admittedly doubtful mineral species. Modern examples of a return to 

 this earlier practice have been offered in late studies of certain varieties 

 of pyrites, feldspars, spodumene and, more recently, jade. 



Recast Analyses of Minerals. 



A few simple illustrations, taken from a series of calculations now at 

 hand, will suffice to show the ease of the long-neglected method and the 

 value of its results. In connection with each analysis, as published, my 

 estimate of the approximate mineral constitution is appended. In con- 

 formity to the description of the mineral, the alumina has been assumed, 

 in these particular examples, as the basis for calculation of the amount 

 either of a chlorite or of an aluminum hydrosilicate, using the theoretical 

 composition which may correspond to the accepted formula of each mineral. 



