142 SYNTHESIS OF DEPSIDES. 



identical with tannic acid, a statement found in text-books even 

 to-day, but which in the Hght of Fischer's work cannot be main- 

 tained. This investig'ation was extended later to a whole series 

 of other phenol carboxylic acids, but unfortimately most of the 

 substances he obtained were amorphous, and hence have been 

 criticised, especially with respect to ttniformity of composition. 

 It is worthy of note, however, that these researches gave rise to 

 the commercial preparation of artificial tanning materials. Fur- 

 ther, Klepl. in 1883, by heating para-hydroxy-benzoic acid, 

 obtained both the above substances, but this simple method is 

 inapplicable in other cases. 



In addition to the depsides, which are pronounced acids, the 

 phenol carboxylic acids can also form neutral anhydrides. 



The introdttction of the methyl carbonato grotip. by tise of 

 the methyl ester of chlor-carbonic acid, is not diffictilt to effect 

 in the case of the meta- and para-hydroxy-benzoic acids, and this 

 is found to be general, if the ])henol grotips are in these positions 

 with respect to the carboxyl groui). At the same time, the 

 number of hydroxyl groups present makes bttt little difference. 

 The hydroxyl groups in proto-catechuic and gallic acids ai"e com- 

 pletely converted with slightly more than the theoretical quan- 

 tity of methyl-chlor carbonate. Btit the case is quite dift'erent, if 

 the hydroxyl group stands in the ortho position to the carboxyl 

 group. The replacement sometimes sticceeds, if excess of the 

 chlor-carbonic ester be used in ac|ueous allcaline solution — e.g., 

 with orsellinic and pyrogallol carboxylic acids — but examples of 

 this are rare. The general method used in this case was first 

 employed by F. Hofmann for the pre])aration of ortho-ethyl-car- 

 bonato-benzoic acid by the action of ethyl chlor carbonate on a 

 mixture of salicylic acid and di-methyl aniline in an indift'erent 

 solvent — e.g., benzene, i'rofessor P^ischer used this method to 

 convert the two hydroxyl groups in gentisinic and ^-resorcylic 

 acids into methyl carbonato groups, from which the chlorides 

 were afterwards ])re])ared ( Bcr., 1909, 215). More difficulty was 

 experienced with the jihloroglucin carboxylic acids, but this was 

 overcome b)' tising large excess of meth}-l chlor-carbonate and 

 di-methyl aniline, and afterwards treating with bicarbonate in 

 acetone solution. 



If the carl)oxyl group is not um'trd directly to the l)enzene 

 ring, e.g., ortho coumaric acid, no difficulty is experienced, and 

 the reaction proceeds well in aqtieous alkaline solution. 



By these methods a large mnuber of the ])henol carboxylic 

 acids have been completel\- converted into the corresjionding 

 methyl carbonato compounds. T'artial replacement ma}^ also 

 occiu-, this being chiefly dependent on the active mass of the 

 reagents tised. but in these cases also it has been found that the 

 para-compound is the one most generally formed, and next the 

 meta derivative. 



The complete reconversion of the methyl carbon.'ito groti]^ to 

 the j)hen(jl group is easily carried out by using excess of cold 



