LARGER CHEMICAL ERRORS. 51 



crystallized salts, such as alum and the vitriols, have been 

 known from the earliest times. 



These bodies are the first indications of definite chem- 

 ical compounds, therefore the very foundation stones of 

 chemistry. 



The power of crystallization also permits us to obtain 

 chemically pure materials for our atomic weight determina- 

 tions. 



We may safely assert that chemical science depends more 

 on this power of crystallization than on any one other; 

 without this power, the very idea of a chemical compound 

 would perhaps not yet have been acquired. 



The native and artificial crystals presented and produced 

 the idea of chemical individuals and compounds first in 

 our mind. 



The most marvelous of all crystals, the diamond, we 

 have found to be the most, if not the only suitable standard 

 of matter for all atomic weight determinations. 



The crystallized carbides, produced by the electric fur- 

 nace of Moissan, were pointed out by us in 1894, as most 

 important materials fit for exact work to connect elements 

 directly with carbon. (True Atomic Weights, pp. 175-176; 



1894.) 



This process has been actually used by Henri Gautier in 

 the Laboratory of Moissan, as will be stated in the chapter 

 on the determination of the atomic weight of boron. 



This method is quite general and will become of great 

 value by directly linking the atomic weight of many elements 

 to that of carbon. 



It is strange that this method was used by Moissan with- 

 out mentioning its origin. 



X. LARGER CHEMICAL ERRORS. 



In modern days we have learnt to produce another class 

 of bodies in almost chemically pure form, namely those 

 bodies which are volatile enough to be distilled. 



Even silver has been distilled in the lime retort by the 

 heat of the oxyhydrogen blowpipe for use in atomic weight 



