WILLIAM PRESCOTT DEXTER. 365 



Rose's laboratory in Berlin that he first went. Even before leaving 

 America, and while still in practice, he revised and published in 1850, 

 in a separate form, the tables for the calculation of analyses which 

 were appended to Rose's work. This entailed a very considerable 

 amount of labor ; for the tables were most thoroughly revised, and 

 nothing but the most perfect accuracy would satisfy him ; and he 

 tells us in the Preface, that " every calculation was performed by 

 myself both by direct division and by the use of logarithms." Also 

 extraordinary care seems to have been taken to avoid all errors of the 

 press. 



After working for some length of time in Berlin, Dexter went to 

 Gottingen and worked in Wohler's laboratory, and from Gottingen 

 he went to Heidelberg. Here he worked for about twelve months in 

 Professor Bunsen's laboratory, and it was during this time that he 

 carried out his most important published work, his determination of 

 the atomic weight of antimony, which was published in 1857. This 

 work was admirably carried out. The writer well remembers the care 

 and thought bestowed upon it, how every possible error he could con- 

 ceive of was considered and eliminated, and the wonderful accuracy 

 with which the manipulation was performed. The method first tried 

 was the simple one of precipitating the antimony by gold ; but this 

 method was found to give inaccurate results, and was abandoned, and 

 the direct oxidation of the metal substituted. Other chemists were 

 working on the same subject at this time, two of them, Dumas and 

 Kessler, obtaining results very nearly agreeing with Dexter's num- 

 ber, 122.34; but Schneider obtained the number 120.3. These 

 researches led to the adoption at that time of the number 122.0. 

 Since then further researches have been made. Cooke in 1880 pub- 

 lished a very elaborate and thorough investigation of the matter. Other 

 investigators and other distinct methods have confirmed the lower num- 

 ber, and 1 20 is now generally adopted as the true atomic weight of 

 antimony. At the same time, it is difficult to see where the error, if 

 there be one, crept into Dexter's experiments. All the papers which 

 Dexter published related to analytical chemistry, some appearing in 

 Poggendorf's Annalen, and some in the Proceedings of the Ameri- 

 can Academy and in the American Journal of Science. His inter- 

 est and his special ability lay in analytical chemistry. He was one of 

 the most perfect manipulators who have ever lived ; the precision and 

 accuracy with which he performed the different processes of analysis 

 were astonishing. Every step was thought out, and then performed as 

 carefully and intelligently as possible. Of late years the discovery of 



