OF ARTS AND SCIENCES. 13 



ing to dull redness. Alkalimetry indicated that 0.32 milligramme 

 should be added to the last weight as a correction for the bromine 

 lost; hence the corrected loss was 0.00018 gramme, or 0.006 per 

 cent. A third trial gave the corrected loss of 3.5 grammes of 

 baric bromide betweeu 185° and a dull red heat as 0.00027 gramme, 

 or 0.008 per cent. Again, 3.4 grammes of a less purr specimen of 

 the salt lost 1.2 milligrammes between 200° and dull redness, of 

 which loss eight tenths of a milligramme was accounted for by the 

 baric carbonate found in the dissolved residue. In Experiment L9, 

 about 3.5 grammes of the salt dried at 260° lost 0.04 milligramme 

 on heating to 340°, and 0.27 milligramme more upon subjection 

 to a red heat. In order to prove that the method of desiccation 

 over sulphuric acid was sufficient for the purpose in hand, this 

 specimen was again heated to 400°, and cooled in a vacuum over 

 phosphoric oxide. After the admission of dry air the crucible and 

 contents were found to have gained a little less than a twentieth of 

 a milligramme. Since seventeen one-hundredths of a milligramme 

 must be added to the last weight of the salt to correct for the 

 amount of alkali found, it is evident that the salt dried at 340° in 

 the first place could not have retained more than 0.005 per cent of 

 water, which could be expelled at a red heat. 



The most severe test of the hygroscopic constancy of baric 

 bromide was obtained by fusion. 17.4841 grammes of baric bro- 

 mide which had been thoroughly dried at a dull red heat were fused 

 in a platinum crucible, and were found to have lost 4.1 milli- 

 grammes during the process. 2.25 cubic centimetres of twentieth 

 normal hydrobromic acid were required to render the solution of 

 the clear cake neutral to phenol phthale'in, and 0.10 cubic centi- 

 metre more made it strongly acid to methyl orange. Thi 

 figures involve a correction of 7.0 milligrammes to the second 

 weight of baric bromide, making it 17.4870 grammes. The 

 cess of this weight over the first (17.4841) is completely accounted 

 for by the knowledge that a slight indeterminable correction" 

 should have been applied to the earlier one, owing to its previous 

 loss of bromine. The crucible was found to have lost 0.20 milli- 

 gramme. 



Again, two grammes and a half of baric bromide heated to con- 

 stant weight at 185° lost 2.11 milligrammes on being Eased in a 

 double crucible. Of this weight all but 0.17 milligramme (0.007 



* From 0.010 to 0.03 per cent. 



