Hydrates of Hydrogen Bromide. 115 



curve obliterating the breaks showed an error 3*5 times 

 larger than the experimental error. Again, a drawing of the 

 results which placed breaks at points other than those above 

 mentioned, namely at 56 and 63 per cent, instead of at 52 

 and 60 per cent., gave an error twice as large as the latter, 

 and 1*7 times larger than the known experimental error. 

 These results, therefore, justify the conclusions drawn from 

 the inspection of the figure. 



Although the breaks thus indicated were of a minor cha- 

 racter only, it was deemed advisable to investigate the freezing- 

 points of the solutions to see whether hydrates corresponding 

 to them could be isolated, and, as a matter of fact, both the 

 hydrates indicated were isolated. Two series of determina- 

 tions (of which the second is the fuller) were [performed in 

 the same manner as in my recent work on the Amines (Chem. 

 Soc. Trans. 1893, p. 141) : these are given in Table II. and 

 illustrated in fig. 2. I need only say in explanation of 

 Fig. 2. — Freezing-points of Solutions of Hydrogen Bromide. 

 F. P, 

 



-40 



-60 



-80 





20 



40 

 Per cent. acid. 



60 



the figure that whenever the freezing-points of solutions rise, 

 attain a maximum, and again fall, their doing so is a proof 

 that the crystallizing substance is a compound of the two sub- 

 stances forming the solutions, and that the surest means of 

 determining the composition of this compound is to ascertain 

 the position of the maximum freezing-point, which must 

 necessarily coincide with the composition of the pure com- 

 pound (see loc. cit. p. 142). Perhaps it may also be well to 



12 



