140 BACTERIA AND THE GUM OF IIAKEA, 



another mucilage a thickening onty was produced. Neutral lead 

 acetate, ferric chloride, copper sulphate, silver nitrate, iodine, 

 tannic acid, Fehling's and Schweitzer's solutions gave no reactions. 

 The hydrates of some of the metals formed insoluble compounds 

 with the gum. Ferric chloride followed by a trace of potassium 

 hydrate gave a transparent yellow clot; barium chloride and 

 alkali produced a mottled transparent and white clot; copper 

 sulphate gave a transparent coagulum. Potassium hydrate 

 thickened the mucilage. These are not quite the reactions either 

 of arabin or of soluble metarabin. A coagulation with neutral 

 1 % ferric chloride is, as for as I know, always obtained with 

 these mucilages. Attempts were made to induce a coagulation. 

 As the gum was acid, the mucilage was neutralised and then 

 dialysed to remove any organic salts that might possibly have 

 prevented coagulation. Still no reaction was obtained. Other 

 methods were tried, such as by making the gum neutral to litmus 

 and precipitating with alcohol, but by no method could a positive 

 reaction with ferric chloride be procured. The probability was 

 that the gum was neither arabin nor metarabin. 



The gum was h3''drolysed by boiling it with 4% sulphuric acid 

 for four hours. The solution contained substances that reduced 

 Fehling's solution, but no recognisable osazones could be obtained. 

 With the same treatment, arabin and metarabin give arabinose 

 and galactose, the osazones of which can be obtained and recog- 

 nised with comparative ease; pararabin is not attacked by that 

 strength of acid. The gum of Hakea, therefore, did not appear 

 to contain any of these substances, so that it differed from the 

 common vegetable gums. Since, however, reducing bodies 

 resulted from the hydrolysis, it seemed advisable to repeat the 

 operation with some alterations. These were the boiling of the 

 acid solution for a shorter time and the use of a smaller amount 

 of acetic acid in the phenylhydrazine mixture. In the course of 

 many recent researches, the proportion of acetic acid had been 

 increased until the mixture contained three parts of glacial 

 acetic acid to one part of phenylhydrazine. The reason for this 

 was that certain indefinite substances that simulated osazones 



