362 



HINRICHS— THE TRUE ATOMIC WEIGHTS [December 2, 



edly : departures (expressed in thousandths of the unit) from the 

 absokite vakies 16 and 108. 



Table of Departures e (in thousandths). 



An extended study of this table would be most instructive, but 

 a brief summary must answer here. 



It is noted that we have given two distinct sets of returns : by 

 intervals of 10 and of 20 thousandths of the departure. In each set 

 we have given only the first two intervals, putting all the others into 

 a third column. The table itself shows the reason for these modes 

 of record in the great predominance of the small departures. • 



This very predominance of the small departures means that the 

 real departures are minute and the actual departures here tabulated 

 as the results of the experimental work done during the century 

 prove the absolute atomic weights to be the real atomic weights of 

 nature. The question of a minute weight of the chemical valence, 

 touched upon in our recent publications, does not come under con- 

 sideration here. 



The table shows that 50 per cent, of all cases give departures 

 below 10 thousandths (o.oi) and 70 per cent, below 20 thousandths 

 (0.02) on all the 656 atomic weight determinations made by all the 

 300 reactions employed. This proves that the determinations have 

 for their limit the absolute atomic weights themselves and that the 

 departures represent mainly the imperfections of the experimental 

 work. 



The relatively few larger departures (only about 12 per cent.) 



