100 JOURNAL OF THE 



What points should decide the choice of our unit? Thev have 

 been ably discussed by Meyer and Seubert,* but I must confess 

 they seem to me somewhat blinded themselves by the partisan- 

 ship of which they accuse their oppouents, and I cannot agree 

 with them in all of their conclusions. 



I would state as the essentials for the unit element : 



1st. That it must be one with which the greatest number of 

 the other elements can be directly compared, thus avoiding the 

 multiplication of error. 



2d. Its own atomic weight must be reasonably small so as not 

 to make too great the higher atomic weights. 



3d. The atomic weights of the other elements gotten by com- 

 parison with it should be, as many of them as possible, integers, 

 or nearly so, rendering calculation easier. In spite of all the 

 tabular aud other aids at the command of the chemist of the 

 present day, calculations with an atomic weight having an awk- 

 ward fraction cause the loss of much time. 



Now, on examining, with a view to these requisites, the two 

 elements proposed as units oxygen alone will be found to answer 

 every requirement. 



Nearly all of the present atomic weights have been determined 

 by the aid of oxygen. 



Few can be directly compared with hydrogen, and this forms 

 the almost insuperable objection to hydrogen as a standard. 



Hydrogen has been used for the past half century for the two last 

 reasons cited among the requisites. As it has the smallest atomic 

 weight, all the others would be above unity if it were taken as the 

 unit. Thus fractional atomic weights were avoided, and again a 

 large number of the other atomic weights compared with it are ap- 

 proximately integers. I am confident that this is mainly made use 

 of as a matter of convenience and of custom at the present day, 

 and that no special weight is attached to the coincidences with 

 whole numbers. It did give rise to a visionary sort of hypothe- 

 sis, first enunciated by the Rev. Mr. Prout in 1815, and hence 



*Berichte der dentpchen Chem. Gesell., XVIII, 1,089. 



