402 BACTERIAL ORIGIN OF GUMS OF ARABIN GROUP, 



no o-um acids and proved to be barium sulphate was filtered from 

 the dilute sulphuric solution. Alcohol did not produce a pre- 

 cipitate in a portion of the filtrate, so that the gum acids appa- 

 rently had been completely hydrolysed by the treatment. The 

 solution was neutralised b}^ boiling with barium carbonate, 

 filtered and the filtrate clarified with aluminium hydrate. A 

 portion of the solution was treated with phenjdhydrazine acetate 

 solution, and a mixed osazone was obtained. When the consti- 

 tuents were separated in the manner indicated in the first portion 

 of this paper, they proved to be arabinosazone (m.p. 158-160°) 

 and galactosazone (m.p. 192-193°). The gum acids therefore 

 contained the arabinan-galactan complex, and the gum belonged 

 to the arabin group. 



The opalescent alcoholic solution which coagulated on the 

 addition of barium hydrate was washed by suspension in alcohol 

 and finally treated with water, in which it formed a semi 

 solution. Dilute sulphuric acid was added until no further 

 white precipitate formed, and the suspension was heated in the 

 autoclave, when a sediment of barium sulphate and a supernatant 

 fluid was obtained. From this, alcohol threw down a small 

 quantity of a contractile precipitate which dissolved easily in 

 water. On testing the solution, which was acid to litmus paper, 

 precipitates were obtained wdth basic lead acetate, ammoniacal 

 lead acetate and barium hydrate. Copper sulphate followed by 

 potassium hj^drate gave a precipitate w^hich contracted but did 

 not darken on boiling. Barium chloride, sodium h3^drate, lead 

 acetate and Fehling's solution gave no precipitates. The reaction 

 with barium hydrate suggests the possibility that the compound 

 may be a pectin substance.* On hydrolysis with sulphuric acid 



* The pectin substances are very indefinite and yield on hydrolysis various 

 sugars. Bourquelot found that gentian pectose hydrolysed to arabinose and 

 doubtful crystals of galactose. Hebert obtained arabinose from the pectic 

 bodies of ripe fruit and bulbs, but as they also yielded mucic acid on oxida- 

 tion, he considered that gaiactan had been present. Bauer hydrolysed apple 

 pectin to xylose and pear pectose to galactose. Herzfeld concluded that 

 parapectic acid contained arabinan and gaiactan. 



