740 ANNUAL REPOKT SMITHSONIAN INSTITUTION, 19 2 8 



the systematic study of the atomic weiglits was begun at Harvard University 

 by the author. 



Observations during the work upon hydrogen under Cooke led 

 him to suspect inaccuracy in the atomic weight of copper, and this 

 subject engaged his energies both before and after his trip to Europe. 

 The thoroughness with which this work was done was character- 

 istic. In order to avoid the danger that a single method might be 

 affected by some constant undetected error, not one but several meth- 

 ods were employed, after each method had been scrutinized with the 

 greatest care. The copper was subjected to the most elaborate purifi- 

 cation, and in order to make sure that copper always possesses the 

 same atomic weight, no matter where it occurs in the earth's crust, 

 specimens from widely different sources were examined. As a result 

 of this work a new value for the atomic weight of copper was ob- 

 tained which has shown no evidence of requiring even slight alter- 

 ation up to the present. The research on copper was followed by 

 similar investigations upon the atomic weights of other common 

 elements, barium, strontium, and zinc being those next attacked. 

 Vp to the time of his death, either with his own hands or with the 

 aid of assistants, Richards redetermined the atomic weights of 24 of 

 the 84 elements which have been isolated in quantity. 



Greater academic responsibilities, as well as an increasing number 

 of research students, early made it out of the question for him to 

 carry on a large amount of experimental work with his own hands, 

 so that in much of his later investigations the laboratory manipula- 

 tion was performed by assistants. The necessity for this is obvious 

 if it is remembered that an expert might spend all his time for a year 

 or even several years in the determination of a single atomic weight. 



In the course of this work many new analytical processes were de- 

 vised and old ones perfected. New methods of purification were 

 invented and new criteria of purity established. Especially Richards 

 appreciated the extreme difficulty, not previously recognized, of free- 

 ing substances, otherwise pure, from the ever-present water, and 

 devised the well-known " bottling apparatus " for inclosing and pre- 

 serving the carefully dried substances in a dry atmosphere j^repara- 

 tory to weighing. The " nephelometer " for comparing and measur- 

 ing traces of solids suspended in liquids was another product of 

 necessity. The importance of taking into account the solubility of 

 " insoluble " substances was pointed out and the great danger of the 

 contamination of precipitates through inclusion and occlusion was 

 emphasized. All these perfections of analytical methods have been 

 of subsequent service not only in the determination of atomic weights 

 but in analytical chemistry in general. 



At the outset of Richards's career the work of the Belgian chemist 

 Stas upon atomic weights was universally accepted as representing 



