OP ARTS AND SCIENCES. 157 



Simultaneously with this reaction, hut in subordinate amount, a 

 direct replacement of the potassium always takes place ; so that an 

 alkylcyanide results as the chief product, but an alkylisocyanide is 

 always noticed as a side product, as has been observed by Dumas, 

 Malaguti, Hofmann, Gautier, and others. 



That the silver salt, AgN=C=, should show a different behavior 

 towards alkyliodides, is also very clear;* first, because silver is 

 less positive than potassium, and consequently the bivalent carbon 

 atom present in both salts must be less reactive in the case of sil- 

 ver cyanide; secondly, a direct replacement of the metal will take 

 place more readily in the case of the silver salt. 



That, however, it may happen, even with the silver salt, that no 

 direct replacement of the metal takes place, is shown by the be- 

 havior of cyanide of silver towards acetyl- and benzoyl-chlorides. 

 In these instances, the isonitriles, CH 3 C0-N=O and C 6 Hr;C0-N=O, 

 are not formed, as Gautier f thought probable in 1869, but 

 CH 3 COON and C 6 H 5 COON are the products % obtained. 



There exists at the present time in organic chemistry an hy- 

 pothesis that a hydrogen atom bound directly to carbon can 

 attain acid properties when one or more negative groups are also 

 bound to the same carbon atom. It is said that the negative group 

 exerts an "acid-making" influence on the hydrogen atom, which 

 extends, however, only to hydrogen bound to the same carbon atom 

 as the negative radical. 



This hypothesis has absolutely no justification, and is entirely 

 illogical. It has already been shown that a methylene group be- 

 tween two carbonyl groups, COCH 2 -CO, has no acid properties 

 whatever. § Furthermore, it is to be observed that in other cases, 

 where it has been assumed that a hydrogen atom bound directly to 

 carbon attains acid properties, the negative group present contains 

 either oxygen or nitrogen, e. g. in nitro methane, 



CH 3 -Nf 



and in HON. Now the groups CI, Br, and F are undoubtedly 

 more negative than the cyanogen group, -ON, and possibly more so 



* Cf. Ann. Cliem. (Liebig), CCLXVI. 137. 



t Loc. cit., XVII. 208. 



} Claisen, Ber. d. chem. Ges., X. 845, XI. 620, 1563. 



§ Ann. Chem. (Liebig), CCLXVI. 67 and 113. 



