LUEDEKING — ANOMALOUS DENSITIES OF FUSED BISMUTH. 293 



The dilatonietric method was used, and as dilatometers ordi- 

 nary tin-case mercurial thermometers. A scale was etched on 

 the stem of each, and this latter then bent at right angles two or 

 three inches from the tip, which had previously been cut off. The 

 metal parts were all discarded. The value of the scale was deter- 

 mined in all parts by the well known method. The entire instru- 

 ment was then weighed after heating to 300° C, the mercury 

 distilled out, and then weighed again empty. Thus the volume 

 of the dilatometer was ascertained at the temperatures of experi- 

 ment, viz. from 250° to 300° C. Also the weight of the glass of 

 the instrument was thus ascertained. It was then refilled, placed 

 in a combustion furnace, and, after the mercury had all been dis- 

 tilled out, the end of the stem, bent at right angles and protruding 

 from the furnace, instantly inserted into molten bismuth. When 

 the bulb was now allowed to cool, the mercury vapor contained 

 therein condensed and molten bismuth was forced into the bulb 

 by atmospheric pressure, filling it wholly. The protruding stem 

 was then seized by means of a pair of tongs and the dilatometer 

 rapidly taken from the furnace and plunged into a mercury bath 

 having a temperature of 300° C. This was allowed to gradually 

 cool, and the readings of the volumes of the bismuth were made 

 just as in an ordinary mercurial thermometer. By means of a 

 thermometer the temperatures of the bath were known during 

 the course of experiment. It will be seen that the whole method 

 consists merely in the construction of a bismuth thermometer out 

 of a mercurial thermometer. The volumes of the instrument be- 

 ing known at each temperature, the various densities of the bis- 

 muth at diftbrent temperatures were thus determinable. Before 

 sufficient skdl was acquired to carry out the experiments satis- 

 factorily months of practice were required, the details to be 

 observed for successful work being quite numerous and un- 

 foreseen. 



The results may be entirely vitiated if the dilatometer is al- 

 lowed to cool to the solidification temperature of bismuth very 

 rapidly. The bismuth under such circumstances rises with great 

 rapidity and suddenness, through the entire length of the capillary 

 stem, immediately before the bulb bursts. The reason of this is 

 easy to find. The bismuth is cooled from the walls of the bulb 

 and deposits crystals on them where they are least thick first. 



