— 23 — 



A hitherto unknown glucoside may therefore be expected, which, 

 when isomerised with baryta yields /^-amygdalin, when completely 

 hydrolysed, d-mandelic acid, and, when partially hydrolysed, sam- 

 bunigrin. 



The isomerisation of amygdalin into w<?-amygdalin indicated in the 

 above table was discovered in 1904 by H. D. Dakin 1 ), who obtained 

 the conversion by treatment with diluted alkalies in the cold. Fresh 

 communications by F. Tutin 2 ) are now available, the upshot of which 

 is that wtf-amygdalin consists of a mixture of two glucosides, viz., 

 amygdalin and neo-amygdalin. 



Tutin prepared /m- amygdalin from amygdalin by allowing it to 

 stand with highly diluted ammonia. The mass having been evaporated 

 to a very thick syrupy consistency, it was at once acetylated, without 

 previous isolation of the tso- amygdalin. The heptaacetylamygdalin, 

 recrystallised from alcohol, had a m. p. of 166 to 167 and a spec, 

 rot. of M D — 37,6° (in chloroform). 



By adding a little alcohol to the crude product of acetylation, 

 diluting the solution with ether, and extracting with water, there was 

 obtained from the ethereal solution a product which crystallised in 

 long, colourless needles and possessed the following constants: — 

 m. p. 1 74 (from alcohol), [a] D — 65,6° (in chloroform), — 57, i° (in 

 acetic ester). The author calls this body hepta-acetyl neo-amygdalin. 



Upon hydrolysis the new glucoside yields d-mandelic acid. 



Up to the present no investigations had been made with the object 

 of determining the process of the hydrolysis of amygdalin by the use 

 of various acids. It was tacitly assumed that they would all have a 

 decomposing action, and that, as is the case with hydrochloric acid, 

 the glucoside, when treated with diluted acids, would yield, in ad- 

 dition to glucose, hydrocyanic acid and benzaldehyde, and, when 

 concentrated acids were employed, mandelic acid and ammonium 

 chloride. Investigations which have lately been made into this sub- 

 ject by W.Walker and V. K. Krieble 3 ) have shown, however, that this 

 supposition is incorrect. When amygdalin was heated for 1% hours 

 with moderately strong concentrated sulphuric acid (2 g. sulphuric acid 

 to 6 g. water) it yielded almost the theoretical quantity of benzalde- 

 hyde cyanohydrin, while, when treated with dilute sulphuric acid and 

 heated for 2 hours at 98°; the decomposition process was found to 

 have only reached the stage of formation of 1-mandelic nitrile glucoside. 

 Concentrated aqueous oxalic acid solution behaved similarly to diluted 



1 ) Journ. chem. Soc. 85 (1904), 15 12. 

 *) Ibidem 95 (1909), 663. 

 3 ) Ibidem 95 (1909), 1369. 



