ELISHA MITCHELL SCIENTIFIC SOCIETY. 107 



reasons for retaining much of this, and the retention or rejection 

 is a point of judgment on which chemists may, and do, legiti- 

 mately differ. An additional sentence of Clarke's reveals the 

 extent of the trouble. " In fact, it is doubtful," he says, " whether 

 any two chemists, working independently, would handle all the 

 data in precisely the same way, or combine them so as to pro- 

 duce exactly the same final results." 



The accompanying tables, giving the calculations of Clarke, 

 Meyer and Seubert (the 0=1(3 table was worked out from their 

 0=1), Van der Plaats, and Ostwald,* show the truth of this. 

 The two last agree more nearly than the others, and yet the va- 

 riations are numerous and sometimes as large as 5 in the case of 

 osmium; 3 in the case of mercury; 1 iu the case of uranium, 

 etc. (nanimity in regard to an atomic weight does not always 

 mean that the weight is correct. Paucity of data sometimes 

 limits the chances for variation. Indium, for instance, and 

 gallium and beryllium have been subjected to few investigations. 



The mathematical side of the question, whether the method 

 of least squares should be adopted, or by what formulae the 

 probable errors should be calculated, and by what the results 

 combined, is, of course, of great importance, and the subject of 

 varying views, but cannot be discussed here. 



*Ostwald's table is put down under the head of 0=16. He gives it as 

 H=l and also 0=16, but the numbers most nearly correspond with the latter. 



