SYNTHESIS SILVER NITRATE. 187 



This little table contains the exact atomic weight of 

 nitrogen, as it results from each single determination made 

 by reducing that determination in the common way by 

 means of the final values of Clarke for the auxiliary elements 

 O and Ag. All determinations made are taken, exactly as 

 recorded by their analytical ratios on pages 62 and 63 of the 

 work of Clarke of 1897. 



Although modern chemists do not look at facts observed, 

 but only at means, and then take these means as facts, they 

 will please not do so in this particular case, but keep each 

 real fact distinctly in view by itself. 



And as the common custom of leaving numbers in a 

 column, does not give the eye a fair chance to see how far 

 these values agree and thus the mouth and the pen may, 

 inadvertently of course, proclaim as facts what is merely error 

 or vain imagination, we shall make sure to avoid such an 

 unhappy result. 



To avoid such a deplorable occurrence of filling the 

 record of science, and then the world, with unmitigated 

 falsehoods and errors, we have taken the trouble to assist 

 the eye of the mind by a simple use of the eye of the head, 

 through plotting the above data. 



In the original drawing, the hundredth of the unit of 

 atomic weight was represented by one inch, the thousandth 

 was, therefore, represented by the tenth of an inch, securing 

 absolute correctness to the third decimal in this drawing, 

 of which a photo-reduction is printed on the lower half of 

 Plate I. 



In order to keep the individual series properly apart, 

 each series was laid down on its own straight line, deter- 

 mined in place by the mean value of that series. 



This somewhat new method of graphical representation 

 gives a faithful and perfectly clear picture of all the facts: 

 the individual observations as abscissae on lines determined 

 in place by their means as ordinates. 



As a matter of course, the locus of the means becomes a 

 line inclined under 45 degrees. 



The results from dried silver nitrate are represented by 

 open circles, those from fused silver nitrate by black circles. 



