BY H. GREIG SMITH. 123 



generally at a wound, and spreads downwards. The microscopi- 

 cal examination of the wood shows the vessels, and especiall}^ the 

 wood parenchyma, filled with a brown gum imbedded in which 

 are mj^riads of bacteria."^ 



So far as I can learn, this disease is not found in Australia, 

 but a gum-tlux does occur. This is not a disease like gummosis, 

 as the health of the plant is not appreciably affected. The gum 

 exudes generally from the surfaces of the branches which have 

 been cut by the pruning knife, and the vines which produce the 

 gum are found in rather damp situations. 



In response to ni}^ enquiries, Mr. Fred. Steward, of Adelaide, 

 S.A., forwarded several portions of vine stems with gum upon 

 the pruned ends. The plants from which the portions were taken 

 had been growing upon a low-lying, rather damp flat. The 

 plants were not unhealthy, and the vigneron could not distin- 

 guish the vines which bore the gum from those in the same 

 locality which yielded none. The sections of the branches had a 

 normal healthy appearance. I also received a small quantity of 

 dry gum which had been picked from the stems. 



The gum consisted of small broken fragments, varying in colour 

 from white to black. They were very dry and brittle, and broke 

 with a glistening fracture. When covered with water the 

 fragments swelled greatly, and the black colour was replaced by 

 a brownish tinge. The gum softened and dissolved in water 

 with extreme slowness. In this connection it must be borne in 

 mind that the gum was collected in November and had been 

 taken from wounds made by the pruning knife in the previous 

 season. The gum had therefore been subjected to many months^ 

 rain, which would probably have washed away any soluble gum 

 that might have been present originally. On boiling with 5 % 

 sulphuric acid the gum acids were hydrolysed and were found to 

 consist of arabinose and galactose, which showed that the gum 

 was of the arabinan-galactan kind. 



* Cent, fur Bakt. 2te Abt. i. 800. 



