52 THE ERRORS OF PRECISION. 



determinations; but we here refer to much more volatile 

 substances which permit their distillation at a temperature 

 readily controlled. 



Such are especially the chlorides and bromides of certain 

 elements not readily subjected to dry way oxidation and 

 reduction. 



We refer here to the use of silicon chloride and bromide 

 by Thorpe and Young and the corresponding compounds of 

 boron by Henri Gautier. 



The silicon bromide used by Thorpe and Young boiled 

 at the fixed temperature of 153 degrees. 



These chlorides react with water, producing the oxide 

 and the corresponding acid. 



The insoluble oxide is generally separated in a weighable 

 form, and the acid may be determined by converting it into 

 the weighable silver compound. 



This mere statement of the process is sufficient to show 

 that it is necessarily much inferior to the ordinary dry way 

 processes. 



As unfortunately great errors have crept into chemistry 

 by the silver process especially when the chloride is con- 

 cerned these apparently very fine methods will be found to 

 be of a secondary value only. 



The silver process here referred to is the comparison 

 with weighed amounts of pure silver, not the volumetric 

 process proper, which latter we shall mention further on. 



Other wet way processes terminating with dry substances 

 and therefore permitting the actual weighing of the final 

 products, have been introduced during the last quarter of a 

 century. 



Good Special Methods. 



One of the most interesting and accurate of these pro- 

 cesses is the conversion of anhydrous borax into sodium 

 chloride by distillation with muriatic acid and methyl 

 alcohol. Sodium chloride is left and weighed. 



It can be objected, that this process is somewhat indirect, 

 since the product does not itself contain the element in 

 question, boron. 



