126 HACTERIAL ORIGIN OF GUMS OF ARABIN GROUP, 



leathery the inner portions in contact with the bark were soft 

 and probably contained living bacteria. 



Accordingl}^, tubes of molten glucose-gelatine were infected 

 with fragments of the soft gum, and some of these were poured 

 into Petri-dishes at once, while others were poured after various 

 periods of incubation at 30°. The colonies that developed from 

 the plates were chiefl}^ those of Bad. acacice. A few other 

 bacteria were obtained, but as these could not be induced to pro- 

 duce gum upon saccharose-potato-agar, or the same with tannin, 

 they were probably adventitious. 



Most of the races of £acf. acacice were similar to the type which 

 I have previously described, but another kind occurred which 

 diiFered from the type in growing as a brownish-yellow mass on 

 saccharose-potato-agar instead of the buff-yellow of the type. 



The natural gum, from which the bacteria had been isolated, 

 when treated with water partly swelled and partly dissolved. 

 The portion which swelled showed rounded faces and corners; 

 most of it dissolved in the course of a month, and doubtless it 

 would all have dissolved in time. Upon hydrolysis with 5 % 

 sulphuric acid the gum acids yielded a solution of reducing sugars 

 which consisted of arabinose anc] galactose. These were identi- 

 fied by means of the osazones which were prepared and purified 

 in the manner already described in the first of this series of 

 papers. 



In view of the slow solubility of the portion of the gum it 

 appeared probable that Bad. metarahinum would occur among 

 the bacteria, but although various methods were tried, and 

 various media were employed, this organism could not be 

 isolated. This failure to obtain the organism, however, does 

 not necessarily imply that it had no part in the production 

 of the natural gum. The separation is attended with difficulty 

 on account of the insoluble nature of the gum which it produces. 

 The slime masses of Bad. metarahinum, instead of dissolving and 

 liberating the bacteria like Bad. acacice, remain intact, and thus 

 a clump of bacteria grows as a single organism. Thus in plate 

 culture Bad. acacict might so outnumber Bad. metarahinum as 



