14 PROCEEDINGS OF THE AMERICAN ACADEMY 



per cent) was accounted for by the amount of alkali found in the 

 dissolved residue (Experiment 13). In Experiment 4, given be- 

 low, the bromide was also fused. Although this sample was not 

 weighed at any lower temperature, it is evident from the amount 

 of silver it required that about the same relation must hold true. 

 It is a necessarj'^ conclusion from these results that baric bromide 

 loses no more water upon fusion than upon being heated to dull 

 redness without fusion. This constancy of hygroscopic condi- 

 tion gives strong ground for the inference that the ignited salt is 

 wholly free from water, and that the salt dried at 180° contains 

 only about seven thousandths of one per cent of the impurity. 

 Moreover, it is very unlikely that water and baric bromide could 

 remain together at a red heat without mutual decomposition. The 

 question of the absolutely anhydrous condition of most substances 

 must necessarily be a matter of inference, because our methods for 

 the determination of a few tenths of a milligramme of water in the 

 presence of a large amount of other material which may be vola- 

 tilized are not sufficiently accurate to furnish direct evidence upon 

 this point. Our knowledge regarding baric bromide is hence as 

 full as it is possible to obtain. 



The specific gravity of baric bromide has been determined by 

 Schiff.* According to his results the crystallized salt is 3.69 times 

 heavier than the same volume of water, while the anhydrous salt is 

 4.23 times heavier. Since it is important in reducing weights to 

 the vacuum standard to know the exact values of these physical 

 constants, new determinations were made. Carefully redistilled 

 dry toluol, in which baric bromide is insoluble, was taken as the 

 liquid to be displaced, and two specific gravity bottles were used. 

 The weight of water filling the first bottle was found upon three 

 trials to be 11.4117, 11.4133, and 11.4120 grammes, these values 

 being corrected to 4° for the expansion of the water, but not cor- 

 rected for the expansion of the glass (24°) nor for the air displaced 

 by the water and weights. An approximate determination of the 

 coefficient of expansion of toluol gave the means of reducing all the 

 weighings with that liquid to the same standard of 24°. Three 

 weighings gave results for the weight of toluol filling the bottle to 

 be 9.8357, 9.8356, and 9.8342 grammes; and 4.4262 grammes of 

 large clear crystals of baric bromide were found to displace 1.0141 



* Liebig's Annalen, CVII. 59; also CVIII. 23. 



