640 SCIENTIFJC KECOKD FOU 18«5. 



On tin; Unit used In Calculatin<j the Atomic Weights (by Lotbar Meyor 

 iiml Karl Seubeit). — The coutroversy arising: immediately alter the pro- 

 l)0.sal of Daltoii's atomic theory, as to the unit upou which the numerical 

 values of the atomic weight should be based has for half a century 

 divided chemists into two schools. While Dalton and, later, Leopokl 

 Gmelin, from theoretical and philosophical considerations, chose the 

 smallest atomic weight, that of hydrogen, as the measure of all the rest, 

 Wollaston and Berzelius chose that of oxygen, jiarily because they did 

 not place so high a value on theoretical views, and partly on the purely 

 practical ground that many elements can be compared directly with 

 oxygen, whereas they can be only indirectly compared with hydrogen. 

 When the Dalton unit, the hydrogen atom, gradually obtained the 

 upper hand, the old controversy appeared to have been laid aside, and 

 consequently it was to be hoped that the recent more exact investiga- 

 tions of the laws which govern the numerical values of the atomic 

 weights would be directed from the same point of view. This hoi)e, 

 however, has unfortunately not been fulfilled, as the old Wollaston- 

 Berzelius unit has lately again come into use in a different — and as we 

 believe — more dangerous form. 



As is well known, J. S. Stas has from his own observation as well as 

 those of others deduced as the most highly probable result that the 

 atomic weight of oxygen is not quite sixteen times as great as that of 

 hydrogen, but on the contrary is about -j^^ of its value less than 1(5 H. 

 That is when H=l, 0=15-96. Many chemists, however, content them- 

 selves with numerical values founded on the more simple ratio of 

 H: = 1 : 16, regarding the difference of ^^ as of very little practical 

 consideration. And the hope that the old controversy over the choice 

 of the unit would disappear upon a recalculation of the atomic weights 

 has unfortunately not been fulfilled. The desire to do away with the 

 unit 0=100 is universal; all chemists prefer referring atomic weights 

 to hydrogen, but they do not agree as to the way in which this is to be 

 done. Some refer all atomic weights to = 16, when H = 1-0023, and 

 others to H=l. In consequence of these different views all the other 

 atomic weights fluctuate to the extent of about ^ per cent, of their 

 value, a very undesirable state of affairs, leading to confusion and per- 

 plexity. The authors maintain that the real cause prompting the 

 adoption of 0=16 is a secret fondness for Front's hypothesis. 



The authors admit that the error introduced by making 0=16 is much 

 smaller than the unavoidable errors of observation so far as inorj/anic 

 compounds are concerned, but they show that in the analysis of organic 

 compounds it is quite otherwise. From a table showing the percentages 

 of hydrogen and carbon in the paraffins containing 30 and 31 carbon 

 atoms and their derived alcohols and acids, it is evident that the vari- 

 ations in the calculated percentages of carbon dioxide reach the tenths 

 of a per cent., a difference often greater than that obtained from two 

 adjax^ent hydrocarbons in a homologous series. 



