Intelligence and Miscellaneous Articles. 551 



With these badly conducting liquids the drop should not exceed 

 h mm. in length. 



Investigated in this way, all dielectric liquids, without exception, 

 exhibit polarization, and, so far as my observations exteud, always 

 in the direction of an increasing surface-tension at the kathode 

 due to polarization. It is, however, prooable that this phenomenon, 

 in the case of benzole for instance, is due to the presence of im- 

 purity ; this seems to follow from the following observations. If 

 the current is closed for a long time, the mobility of the drop is 

 considerably diminished. In like manner, when the apparatus is 

 left for some time at rest, the conductivity of the dielectric liquid 

 is considerably diminished. 



If this idea is correct it ought to be possible, by carefully puri- 

 fying the substance, to dimiuish as much as possible the peptization 

 and therefore also the conductivity. I used for my experiments 

 benzole (not benzine, for this is no well-defined body), which Dr. 

 Haussermann was good enough to prepare for me in the greatest 

 state of purity. "When this was examined in the electrometer in 

 the manner described, it showed, like a commercial specimen, 

 strong polarization, notwithstanding that the electrometer had 

 previously been carefully purz'fied and dried, and mercury used 

 which had been repeatedly distilled. Only when, in addition to 

 the other precautions, the mercury was filled hot (about 120°) into 

 the hot apparatus, did the benzole show no polarization, even with 

 high electromotive forces. Benzole in its greatest purity appears 

 therefore to be a perfect insulator for electricity ; the conductivity 

 observed depends on impurities and particularly moisture, which 

 in these experiments may have been on the surface of the mercury. 

 Those methods of determining the dielectric constants of dielectric 

 liquids, the result of which is influenced by a possible conductivity 

 oi ! the liquid, must therefore lead to wrong results, for in the 

 ordinary arrangement and with the comparatively large masses 

 and surfaces which come into play, it must be difficult to prevent 

 the impurities and the presence of moisture. 



As regards sensitiveness the method is equal, if not superior to 

 a galvanometric method with a very sensitive astatic Thomson's 

 galvanometer. — Wiedemann's Annalen, No. 11, 1893. 



ON THE EMISSION OF HEATED GASES. BY F. PASCHEN. 



The chief results of his investigation of this subject are thus 

 summarized by the author : — 



Gases can in consequence of their temperature emit a discon- 

 tinuous spectrum ; hence the opposite view T of M. Pringsheim 

 fails. This emission has been investigated for 0O 2 and aqueous 

 vapours at different temperatures. A displacement of the in- 

 tensity within the maxima of emission was observed with decrease 

 of temperature. This fact could be demonstrated but not ex- 

 plained. 



