248 PROCEEDINGS OF THE AMERICAN ACADEMY. 



of acids of phosphorus never became high in this U-tube, but it was 

 evident that a single such tube was not enough to remove the bromides 

 of phosphorus completely. Consequently a second U-tube, filled with 

 beads moistened with hydrobromic acid, was inserted, and pains was 

 taken that no very considerable concentration of acids of phosphorus 

 should ever exist in the hydrobromic acid generating flask. After this 

 modification of the apparatus had been made, the boat remained absolutely 

 unchanged in appearance and weight. 



These phenomena became very evident in the present work because 

 such large quantities of hydrobromic acid were necessary for the synthesis 

 and sublimation of ferrous bromide. Also, since the bromide was con- 

 densed from an atmosphere of the gas, the conditions were those capable 

 of causing the retention of a maximum amount of phosphorus. In the 

 cases of the cobaltous and nickelous bromides, on the other hand, syn- 

 thesis was effected by means of bromine vapor, very little hydrobromic 

 acid being used in their preparation. It was only during the drying of 

 the salts that they came in contact with hydrobromic acid for any con- 

 siderable length of time, and even then the contamination could have 

 been only superficial. As a matter of fact, it was found impossible at 

 the time to detect the slightest trace of phosphorus in either of the 

 two bromides. 



Considerable difficulty was experienced in obtaining ferrous bromide 

 which would give a clear solution in water. The earlier samples con- 

 tained, besides silica from the new porcelain tubes, a slight black residue 

 which became red upon ignition, and which may have been the magnetic 

 oxide of iron. The amount of this residue was very much diminished 

 by a lavish expenditure of hydrobromic acid during sublimation. These 

 residues were filtered upon tiny filter papers, ignited, and weighed. In 

 none of the samples used for analysis did the weight of the insoluble 

 residue amount to more than one thirtieth of a per cent of the weight of 

 the salt. Even in the best material the insoluble residues contained 

 traces of phosphoric acid. In Analysis 2 (p. 252) the phosphoric acid 

 in the residue was determined by precipitation with amnionic molybdate 

 to be 0.0001 gram. It was easily shown, however, that all the phos- 

 phoric acid was in the insoluble residue, for in Analysis 4 the filtrate 

 from the silver bromide was evaporated to small bulk and treated with 

 amnionic molybdate, without the formation of a visible amount of pre- 

 cipitate. Evidently no error was introduced into the analysis by the 

 presence of the phosphoric acid. 



The determination of the bromine in the salt offered certain difficulties 



