Hydrates of Hydrogen Bromide. 113 



Owing to the large gaseous tension of very strong solutions 

 of the acid, it is impossible to make them up to definite 

 strengths by diluting the strongest acid with weighed quan- 

 tities of water. The strength of each solution had to be 

 determined separately after its density had been taken, for 

 which purpose the contents of the density-bottle were emptied 

 into a large weight of water, and portions of the solution 

 obtained titrated. Three or four titrations of each solution 

 were made, and from the variation in the values obtained the 

 average mean error of the results was deduced to be 0*0466 

 per cent. HBr, corresponding to a mean error of 0*00045 in 

 the densities. This error is, of course, far larger than the 

 error in the density-determinations themselves ; indeed the 

 latter is practically negligible in comparison with it. From 

 44 per cent, downwards the results are more accurate, the 

 solutions having been made up to the required strength by 

 weighing. 



The very strong solutions of the acid cannot be made and 

 kept without developing a considerable amount of free bro- 

 mine. This renders the titrations somewhat uncertain, and 

 has the effect of making the densities appear too high ; it 

 may also affect the apparent position of any changes of cur- 

 vature which exist, though in what direction it is impossible 

 to say, for when the strength of the solution is determined 

 by titration, any bromine which is present will be reckoned as 

 water, and too low a value for the relation of acid water really 

 present may thus be obtained ; but, on the other hand, the 

 bromine may really be dissolved in and combined with some 

 of the acid, leaving a smaller amount than the total found by 

 titration operative, or combined with the water present. We 

 must, therefore, not expect the position of any breaks to coin- 

 cide very exactly with definite molecular proportions as based 

 on the results of titration. The weaker solutions might have 

 been obtained nearly free from bromine ; but as the main 

 consideration in the present determinations was concordance 

 of consecutive results, and not their absolute accuracy, it was 

 necessary to use similar solutions throughout the series. 



The results when plotted out form a figure illustrated by 

 A B (fig. 1) . This figure, unlike that representing the den- 

 sities of hydrogen-chloride solutions, exhibits very considerable 

 curvature as far as about 47 per cent., and afterwards is but 

 slightly curved ; the position of this general change, not 

 necessarily a definite change of curvature, corresponds with 

 the strength at which the solutions cease to give off water on 

 exposure to air and begin to give off acid (Roscoe, Quart. 

 Journ. Chem. Soc. xiii. p. 157). There seem to be indications 



Phil. Mag. S. 5. Vol. 36. No. 218. July 1893. I 



