192 E. von Aubel on the Influence of Magnetism 



were studying, notably tbe excellent memoirs of Messrs. 

 Goldhammer*, A. von Ettingshausenf, and W. NernstJ. 



Quite recently we saw, in the Wiener Anzeiger, 1887, 

 p. 222, that Messrs. von Ettingshausen and Nernst had made 

 a communication to the Academy of Sciences at Vienna, 

 having the title " Ueber das thermische und galvanische 

 Verhalten einiger Wismuth-Zinn-Legirungcn im magnet- 

 ischen Felde." This paper is not yet published, as far as we 

 know ; so that we cannot take account of it in what follows. 



Our researches are far from being finished. They will be 

 continued as soon as our other occupations will permit. 



I. The Preparation of Rods of Bismuth. 

 We have studied bismuth under three molecular states : — 



(1) Melted and cooled slowly, 



(2) Melted and rapidly cooled or tempered, 



(3) Compressed. 



Let us examine the preparation of rods of bismuth in these 

 three cases. 



1. In order to prepare bismuth wires slowly cooled, 

 capillary glass tubes were procured, to which were joined 

 at both ends, at right angles to the capillary part, glass tubes 

 of fairly large diameter. These capillary tubes were heated 

 in a sand-bath, and some grains of bismuth were introduced 

 into one of the side branches. When the metal was melted, 

 it ran down by its own weight into the capillary tube. The 

 capillary tubes were then left to cool very slowly ; they were 

 taken from the sand-bath only when it was quite cold. It 

 often happens that, owing to the unequal expansions pro- 

 duced during the cooling of the bismuth and the glass, the 

 tubes are broken ; for this reason we used capillary tubes 

 with very thin walls. 



Working in this way, any formation of oxide in the 

 capillary part is entirely avoided ; and this part alone must 

 be used in the experiments. 



We think that this way of preparing bismuth wires is much 

 more convenient than the method by exhaustion employed by 

 others§, and which cannot furnish results entirely comparable, 



* We have given in the Bibliography only the list of works relating to 

 bismuth. 



t Vide the Bibliography. 



X Annalen der Physik, 1887, No. 86. 



§ See notably the memoir of Leduc, Journal de Physique [2], iii. 

 p. 3G3. 



