and Electrolysis of Chemical Compounds. 383 



elements, only 2°; and this may arise from a trace of water, 

 since the preparation of the first compound was effected by a 

 freezing-mixture, and the second, although it forms a well- 

 crystallized substance, is extremely hygroscopic : it is there- 

 fore with difficulty protected from the aqueous vapour of the 

 air. This is more successfully accomplished with carbonic 

 anhydride, which then exhibits no conduction at all *. 



Of the chemical reactions of these bodies little is known. 

 With the aid of a little heat, various metals act reducingly on 

 them. Gore says that liquefied carbonic anhydride exerts no 

 chemical action, has but very slight solvent power, and is re- 

 duced only by the alkali-metals. 



The anhydrides which contain metals belong (so far as is 

 known) to the good conductors, with the exception of osmic 

 anhydride. This compound, which fuses at 40° C. is not de- 

 composed by a very powerful galvanic current (80° B. e.), 

 although its chemical properties are similar to those of the 

 three other anhydrides, which were split up by a battery of 

 1 6 Bunsen elements t ; and it parts with its oxygen as readily 

 as these do to oxidable substances. In a similar manner it 

 is seen that facility of chemical decomposition is not accompa- 

 nied by high conductivity in hyponitric acid (N0 5 ), which, 

 however, is not properly an anhydride, and therefore is not 

 included in the above Table. This gas was liquefied by a 

 freezing-mixture, and was perfectly pure. In this condition 

 it was completely non-conducting to the current from 80 

 Bunsen elements; and with the electrodes only one millim. 

 apart an induction spark of 15 millims. length would not pass. 

 Yet, according to Hittorf s theory, this compound should be 

 an electrolyte, since with water or hydrate of potassium it 

 suffers decomposition by elective affinity. In the liquid state 

 it acts on metals and metallic oxides, with which it forms 

 nitrates, at the same time liberating nitrous oxide f. As in 

 the case of hydrofluoric acid, so here the pressure was only that 

 of the atmosphere, and the metals (as lead, mercury, potas- 

 sium, and sodium) reacted at a very low temperature, although 

 somewhat slowly. The chemical properties appear therefore 

 not to be much modified by the liquid state alone. 



The anhydrides of the organic acids are, within the limits 



* Gore, Phil. Trans. 1861, pt. i. p. 83. When the electrodes were 2 

 millims. apart the current of 40 Smee elements did not pass through ; and 

 just as little did the discharge of an induction-coil when only 0*2 millim. 

 of liquid separated the electrodes. 



t Their electrolysis has been effected by Buff: see Pogg. Ann. vol. ex. 

 p. 267. 



| Richard Miiller, Liebig's Annalen, vol. exxii. p. 1 (1859). 



