BY R. GREIG SMITH. 249 



made. Two months afterwards the tree was examined, and small 

 pellets of yellowish gum were found at the infected places. In 

 this case the gum had not migrated down the stem. The granules 

 of gum simply swelled with water and did not dissolve, thus 

 showing it to be metarabin. It is possible, and I think it quite 

 probable, that as the gum issued from the wounds it consisted of 

 a mixture of arabin and metarabin, and that occasional rains had 

 dissolved the former and washed it down the trunk of the tree. 

 Control trees show no appearance of gum, and, therefore, we 

 must conclude that the exudate was caused by the infected 

 bacteria. 



Bact metarahinuyn, a variety of Bad. acacice. — When we 

 remember that Bact. acacice and Bact. metarabinum virtually 

 differ in the production of arabin and metarabin respectively,* 

 it is significant that Bact. acacice should, when infected into the 

 peach, produce a gum-flux of insoluble metarabin instead of the 

 soluble arabin. The gums of the Rosace?e are practically always 

 metarabin (or cerasin, as it is frequently called), and this experi- 

 ment shows that the host tree can alter the gum-forming faculty. 

 This it must do either by modifying the gum after its formation 

 by Bact. acacice or by modifying the bacterium, so that it becomes 

 a metarabin producer, i.e., Bact. metarabinum. To gain some 

 information concerning this, I examined the tissues of a branch 

 of the infected tree. The branch sprang from a place between 

 the infected areas of the stem. In the plates of glucose gela- 

 tine that were prepared there developed colonies of Bact. acacice, 

 and upon one of the plates I obtained an impure colony of Bact. 

 metarabinum growing upon an imbedded fragment of bark. I 

 have already! discussed the difficulty experienced in isolating Bact. 



* With the exception of the gum, the bacteria form the same products 

 during the fermentation of saccharose, and they are morphologically very 

 similar. If the gums had an equal solubility all the cultural characters 

 would probably be identical. In short they appear to be races of one 

 organism producing different kinds of gum. But the gums so modify the 

 characters of the cultures that the races appear to be species. Believing 

 this to be the case, I subcultivated both bacteria at 17*^ and at 30'^ for four 

 months to see if the characters would approximate, but no change occurred, 

 t These Proceedings xxvii. (1903), 126. 



17 



