OF ARTS AND SCIENCES. 7 



During the latter part of the investigation the balance was kept 

 in a small room built entirely inside of the main laboratory. The 

 absence of outside windows in the small room caused a notable ab- 

 sence of air currents and rapid changes of temperature, while its 

 glass walls supplied plenty of light. 



It is almost needless to state that, while the weights of the appa- 

 , ratus were not reduced to the vacuum standard, — on account of the 

 method of weighing, which rendered such reduction unnecessary, — 

 the weight of every substance used was corrected in the manner 

 shown above for the difference between the weight of air displaced 

 by it and that displaced by the corresponding brass weights. 

 Where the specific gravity of the substance was not already 

 accurately Known, it was carefully determined for this purpose. . 



The Spectroscopic Detection of Calcium and Strontium 

 IN THE Presence of Barium. 



In the course of the search for a typical barium salt it became im- 

 portant to determine how small an amount of calcium and strontium 

 could be detected in the presence of large amounts of barium. The 

 most sensitive method is naturally the spectroscopic one, but no 

 literature giving the degree of sensibility seemed to be at hand. 



The first phase of the problem to be investigated was the deter- 

 mination of the amount of calcium and strontium which could be 

 detected in the absence of barium. Hence a standard solution of 

 calcium and strontium was prepared containing 0.8 milligramme 

 of each metal to the cubic centimetre. This solution was succes- 

 sively diluted and tested by means of a well made single prism 

 spectroscope with an adjustable slit. A drop of the solution was 

 supported npan a coil of wire containing 0.018 cubic centimetre, 

 similar to that suggested by Truchot* and so ably used by Gooch 

 and Hart.f No attempt at quantitative analysis was made, the 

 present problem being merely the determination of the limit of 

 visibility. 



* Compt. Bend., LXXVIII. 1022. 



t Atn. Journ. of Sci., [3.] XLll. 448. The writer is much indebted to this 

 paper for valuable hints. 



