144 S^'N'J-IIKSIS OP DKl'SIDIiS. 



In the case of the (U- ami tri-h\(h"()xv henzoic acids, con- 

 taining two and three hydroxy grottps, isomeric methyl carhonato 

 depsides result, as also complicated derivatives, which are difhciilt 

 to separate and ])urify. So far one case only has been fully 

 worked out, the coupling of di-methyl carhonato orsellinic acid 

 chloride with orsellinic acid necessar^■ for the synthesis of 

 lecanoric acid. 



Digallic, di-])rotocatechuic acids, as well as a large number of 

 other di-de])sides, have been prepared by this method, and their 

 constitution thus clearly established. Many of these substances, 

 as has been already mentioned, show the reactions of the tannins 

 in greater or lesser degree. 



On the subject of these clilorides. Professor Fischer further 

 remarks that their apjjlicability for synthetic pttrj^oses is by 

 no means exhattsted, and that they will yet prove of great service 

 in cases where benzoyl chloride has hitherto been found of ser- 

 vice, and where the ])henol gnnijjs may be easily restored by 

 gentle saponification. Me further recommends the tise of the 

 chloride of methyl car])onato ferttlic acid as suitable for the syn- 

 thesis of curcttmin, which has not yet been worked out. 



The reconversion of the methyl carhonato grotip to phenolic 

 .hydroxyl is possible by using cold dilute alkali or acjueous 

 ammonia at 20° C, the reaction being generally completed in half 

 to three-f|uarters of an hour, with normal or semi-normal solu- 

 tions. If alkali be tised. sufficient is added to netitralise the car- 

 boxyl group, and then two molecules NaOH for each methyl 

 carhonato grou]). If the weaker ammonia be em])loyed, a large 

 excess is reqtiired. 



The di-dei)sides hitherto investigated are crystalline stib- 

 stances, difficultly soluble in cold water, but soluble in bicar- 

 bonate solution, acid in reaction, and melt with dec()m])osition. 

 Like the simple phenol carboxylic acids, they give colotirations 

 with ferric chloride. They react with diazo methane, forming 

 coiupletely methylated comjjounds. which crystallise extremely 

 well, and melt at a lower temperature than the de])sides without 

 decom])osition. Owing to these properties the\' have been usefttl 

 for the i)uriK)ses of identification. 



The first i)ai)er directly concerned with tannin a])])eared in 

 the Bcrichic in 1912, and was entitled " Tannin, and the Syn- 

 thesis of Siniilar Substances." It begins with an historical intro- 

 duction discussing briell}' the various researches i)ul)lished up to 

 that time. These had not resulted in the establishment of a 

 defmite com])osition or fornuila for tannin which would defy 

 criticism or investigation, but certain facts had been ascertained 

 which Professor Mscher thought it necessary to contirm or 

 reject as a restilt of his own investigations. 



Under the generic n.ime of tannin, a large number of vege- 

 table substances from different soiures have been included, which 

 have been classified at various times in various ways, according 

 to the ])oint of view chosen. The experiments undertaken in this 



