and Electrolysis of Chemical Compounds* 379 



It follows from these experiments that most of the hydrogen 

 acids are extraordinarily bad conductors of the galvanic cur- 

 rent. This behaviour is the more remarkable as the compounds 

 here present are, when dissolved in water (likewise a very 

 bad conductor), some of the best conductors and electrolytes. 

 For this very reason, because a slight trace of water is sufficient 

 to much increase the conductivity, the negative result may be 

 regarded as a proof of the adequate purity of the condensed 

 gases*. The only exception is hydrocyanic acid, it being 

 decomposed by the current of only 3 or 5 small Bunsen ele- 

 ments. I endeavoured to condense the gas perfectly dry by 

 decomposing dry cyanide of mercury with dry sulphuretted 

 hydrogen and conducting the gas formed over phosphoric 

 anhydride before it entered the condensation-tube. At the 

 negative pole hydrogen was evolved ; and at the positive pole 

 platinum cyanide was probably formed, which coloured the 

 liquid yellow. Moreover this compound is readily decom- 

 posed chemically ; it is even decomposed by light, with forma- 

 tion of ammonia. 



These bodies retain their insulating properties when sub- 

 mitted to electricity of high tension ; for only the more 

 powerful spark-discharges can take place in them when the 

 distance between the electrodes is from 2 to 4 millims. The 

 conduction of induction-currents is in this case complicated, 

 since it is difficult to decide whether a merely mechanical 

 discharge takes place, occasioning, by the production of heat, 

 a sort of dissociation of the compound, or whether it is ac- 

 companied by a proper electrolytic action. 



With hydrochloric acid f, even when moderate induction- 

 currents were employed, an effervescence was observed at the 

 electrodes. In order to determine the cause of this, the in- 

 duction-currents (spark-length 15 millims. in air) were 

 directed through the liquid acid during four hours ; but no 

 spark-discharge took place. The part of the tube containing 



* This influence of slight admixtures is evident, for instance, from an 

 observation made by Becquerel, who was able to decompose alcohol with 

 a battery of 50 elements if only ^— of potassium was added. Despretz 

 found the electrolysis of water with 2 -^Vo °f sulphuric acid different from 

 that of distilled water. The same fact has also been pointed out by 

 Kohlrausch (Pogg. Ann. Ergzbd. viii. p. 4). 



t The conductivity of chloride and fluoride of hydrogen in the liquid 

 state has also been investigated by Gore (Proc. Royal Soc. vol. xvii. p. 

 256). With a pile of 10 Sniee's elements and with the electrodes sepa- 

 rated by a distance of 0*16 millim. the galvanometer showed a 3mall 

 deflection ; likewise with the induction-coil conduction was perceptible 

 accompanied by chemical action. Gore, however, thinks that the liquefied 

 gas was not perfectly pure. Hydrofluoric acid, with gold electrodes, ex- 

 hibited almost no conduction. 



