1 82 STASIAN FOLLY AND FRAUD. 



3 thousandths, will give us for the probable error of a single 

 determination 8 thousandths, that is 0.008. 



Since the mean of Stas is 14.041, half of all observed 

 values should lie within the range 0.008 below and above this 

 mean, that is between 14.033 and 14.049; let us say, between 

 14.03 and 14.05. See bottom p. n, and pp. 16-17. 



For the " fused" nitrogen atoms here considered, only 

 No. 6 falls within this limit the other six determinations 

 are far above it, being located between 14.06 and 14.08. 



For the "dried " atoms, a similar state of facts would result. 



But why will common laborants, who like the janitors 

 of a chemical laboratory, know of no chemistry beyond the 

 mixing of liquids and the ignition of solids, with more or 

 less of stink and fumes, meddle with mathematical processes 

 they do not understand? 



It is like playing with new firearms they may find them 

 loaded when least suspecting such a thing. 



They ought to be more careful, hereafter. 



II. THE ATOMIC WEIGHT OF NITROGEN BY CHEMICAL MEANS. 



Really it is not necessary to enter upon the purely chem- 

 ical determinations of the atomic weight of nitrogen, after 

 the perfectly unquestionable results obtained by density 

 determinations of Lord Rayleigh and the preceding little 

 Note on Stas' Syntheses of silver nitrate and his challenge. 



But the muddled state of the chemical mind, produced 

 by the pretenses of Stas and diffused by the high chemical 

 and academic endorsements of Stas, which have made the 

 expression of any doubt about the Stas values a heresy, 

 compel us to enter upon this chemical part once more. 



I must be permitted to insist that a demonstrated fact, 

 such as the atomic weight of nitrogen by density determina- 

 tions, is to be received as such, and cannot be suppressed. 

 It must be accepted as a finality. 



Anything in conflict with such a fact proves itself to be 

 in error. If the atomic weight of nitrogen obtained by 

 strictly chemical means differs from 14, we can only look for 

 causes of error in these methods of determinations. 



But as a still further concession to the deplorable lack of 



