148 BACTERIA AND THE GUM OF II AREA. 



Barium hydrate gave a slight stringy precipitate. Coagula 

 were not obtained with Fehling's solution or with copper sulphate, 

 although when the latter was followed by a trace of alkali a clot 

 was formed. These are practically the reactions of arabin, and 

 especially of dissolved metarabin. This is exactly what would 

 have been expected from pectin, and supports the idea that the 

 natural gum is that substance. 



Unfortunately for this supposition, the hydrolysed products 

 of the alkali-treated gum did not contain either arabinose or 

 galactose, and from this we must conclude that the gum is not 

 pectin. In testing this point, the digested guin was boiled for 

 two hours with 1 % potassium hydrate, neutralised and precipi- 

 tated with alcohol. The flocculent precipitate was boiled for 

 three hours with 5 % sulphuric acid, neutralised and treated with 

 alcohol, when a very small precipitate was thrown out. The 

 solution, after elimination of the alcohol, was treated with 

 phenylhydrazine acetate and the resulting osazone-like bodies 

 purified with ether. They had much the same indefinite character 

 of the substances previously obtained, and melted at about 120°, 

 140° and 175°. Neither arabinosazone nor galactosazone could 

 be detected. 



The conclusions to which this research has led are as follows: — 



1. The gum of Hakea saligna is neither arabin, metarabin, nor 

 pararabin. The hydrolytic products consist of reducing bodies 

 that yield indefinite osazones and are probably akin to the fur- 

 furoids of Cross, Bevan and Smith. It is not pectin, although it 

 approaches this substance in some respects. 



2. Of the bacteria occurring in the tissues of the plant, the 

 most probable producer of the gum is one intermediate between 

 Bact. acacice and its variety, JJact. 7netarabinum, but as we do 

 not yet know that the host-plant can alter a gum once formed 

 by a bacterium, it cannot be said that the gum is produced by 

 this micro-organism. 



3. Bacteria that produce galactan gums which behave to 

 reagents like arabin are not uncommon. A second is described 

 under the name of Bacillus pseudarabiyius ii., n.sp. 



