ENDOTHIA PAEASITICA AND RELATED SPECIES. 



65 



TABLE VI. Inoculations with Endothia ftuens mississippiensis on Acer and 



Carya. 



As in Endothia fluens the growth was confined to the injured tis- 

 sues, and there was no evidence of parasitism. 



ENDOTHIA TROPICALIS. 



The material of Endothia tropicalis from which the writers se- 

 cured their cultures, was collected by T. Fetch in Ceylon. As the 

 species of Endothia in the Northern Hemisphere are chiefly on 

 members of the Fagaceae, Fetch's statements with regard to hosts 

 are of considerable interest. In a letter of March 6, 1914, he writes : 



We have no Fagacese native in the island. We have introduced various 

 species of Quercus and Castanea, but subsequent to Thwaite's discovery of 

 this fungus. I do not think there can be any doubt that the fungus is native 

 to Ceylon * * * 



Of the speciments now sent * * * those in the packet * * * are 

 from a tree which was producing shoots from the base. This tree is Elaeocar- 

 pus glandulifer Mast. From the bark and habit, I believe that all my " finds " 

 of Endothia have been on this species. 



In the accounts of the American ckestnut disease," I 'notice that several 

 authors speak of " cankers," and give their rate of growth. I never see 

 "cankers" (Krebs) on the Ceylon trees. The bark appears to die regu- 

 larly and smoothly from above downward, and is quite unbroken except for 

 the minute cracks through which the stromata emerge. 



Inoculations. As already noted, ascospores of Endothia tropi- 

 calis resemble those of E. parasitica even more closely than do those 

 of E. fluens. This fact, together with its similarity on culture media 

 and its oriental origin, led the writers to fear possible parasitic 

 tendencies. 



Inoculation experiments were accordingly made only on the chest- 

 nut and under carefully guarded conditions. In all, about 30 

 inoculations were made on 2-inch chestnut sprouts, using the methods 

 described for other species. 



Of 25 inoculations made in May and June, practically all had de- 

 veloped a few pycnidial stromata by October 20. These stromata 

 were a somewhat brighter orange than those of E. fluens or E. fluens 

 mississippiensis, and the spores when cultured produced typical E. 

 tropicalis. In no case, however, was there any evidence of parasitism. 

 43737 Bull. 38017 5 



