LUEDEKIXG ANOMALOUS DENSITIES OF FUSED BISMUTH. 297 



eters used by Vicentini were not sensitive enough for solution of 

 the question concerning anomalous expansion of fused bismuth. 



I now proceed to compare his co-efficient of expans'on of fused 

 bismuth with that obtained in my experiments. It is nearly three 

 times as great. In the extensive series of experiments made by 

 Vicentini, his results agree very well with one another. I pre- 

 sume, therefore, that the difference is due to a constant error in 

 the method, which I think is to be found in his mode of filling 

 the dilatometers. We know how tenaciously air adheres to glass 

 vessels, and how difficult it is to get rid of it. Even that dila- 

 tometer par excellence, the mercurial thermometer, contains air 

 in appreciable quantities in most cases. Vicentini in simply 

 filling his dilatometers with bismuth, and without resorting to 

 the known methods of freeing them from air, must have quite 

 a constant error in his absolute results. His co-efficient would 

 be too high from this cause, whilst mine must be comparatively 

 free from such influences. 



Chemical Laboratory Washington University. 



