26 RESEARCHES UPON ATOMIC WEIGHTS. 



varied by sometimes adding the silver nitrate to the cadmium bromide (analyses 

 1,2, 3, 6, and 7), and sometimes adding the bromide to the silver nitrate 

 (analyses 4, 5, and 8) . And in the third place the solutions were allowed to stand 

 varying periods before the titration was completed, so that occluded substances 

 might have opportunity to be dissolved. Analysis i, in which the largest quan- 

 tity of bromide was employed, over 11 gm., which is to be expected to give the 

 most marked evidences of occlusion, was not tested for 5 days after precipita- 

 tion, and the titration was completed 8 days later. In the other analyses the 

 period between precipitation and the completion of the titration varied from 7 

 days in analysis 4 to 3 days in analysis 8. Furthermore, in some cases, after the 

 end-point had been reached, the solutions were allowed to stand some days 

 longer with occasional testing. No change in end-point with standing was ob- 

 served. In spite of these differences in the method of procedure, the variations 

 in the final results do not exceed the experimental error to be expected, except 

 in the case of analyses 4 and 12. Evidently, occlusion of any sort must have 

 been very shght if it existed at all. Analyses 4 and 12, performed with the 

 same portion of bromide, differ so markedly from the others that, although no 

 reason for the difference is known, they are rejected in computing the final 

 average. 



The gold-plated brass weights were carefully standardized to hundredths of 

 a milUgram. Vacuum corrections of -|-o.oooo86 for cadmium bromide,^ of 

 +0.000041 for silver bromide and of —0.000031 for silver were applied. All 

 weighings were made by substitution with counterpoises as nearly like the ob- 

 jects to be weighed as possible. 



The analytical work was performed wholly by Dr. Hines. (See table on 

 opposite page.) 



RESULTS AND DISCUSSION. 



The ratios of silver used to silver bromide obtained in the same analysis afford 

 sufficient proof of the purity of the bromine and silver, as well as confirmatory 

 evidence of the absence of appreciable occlusion by the silver bromide. 



Ag : AgBr. 



Analyses i and 9 57.4446 



2 " 10 57.4466 



3 " II 57-4438 



4 " 12 57-4438 



5 " 13 57-4423 



6 " 14 57-4440 



7 " IS 57-4430 



8 " 16 57-4431 



Average, 57-4439 



* The specific gravities of cadmium and silver bromides have recently been found to be 

 5.192 and 6.473 respectively. Baxter and Hines: Amer. Chem. Jour., 31 > 220 (1904). 



As in the case of cadmium chloride, slight changes in the vacuum corrections of cadmium 

 and silver bromides have been made from the values employed in the original publication 

 of this paper, owing to more exact knowledge of the density of the weights. 



