H. WoJi.MALl) 187 



the number of trusses treated (i.e. one flower of each inoculated), and 

 in the other is the number of successful inoculations resulting in each 

 case in the death of the whole truss. 



The blossoms of the Prince Bismarck trees were the first to ex])and 

 and the earliest of these were inoculated immediately after opening, 

 some of them on the first day on which the stigmas were exposed. The 

 flowers of the other varieties were as a rule rather older when inoculated 

 and this suggests (judging from results as indicated in the table) that 

 the flowers are most susceptible to infection immediately after they 

 open. None of the untreated trusses showed any signs of wilting 

 throughout the season except in those cases where the fungus had 

 extended from the inoculated flowers so far as to cut off supplies to 

 those parts not directly infected. 



The results obtained from these experiments, in those instances 

 where inoculation of a flower was followed by the wilting of the remaining 

 flowers and also the leaves of the spur, are given below in detail to 

 illustrate the progress of the disease from the time conidia were placed 

 in the flowers. 



All the trees were removed from the greenhouse about the middle of 

 May and were left in the open throughout the winter. 



Variety Prince Bisnuirck. 



(a) Flower of a spur 1 cm. in length inoculated April 4 by placing 

 conidia on the stigmas. On April 11 the styles were dark brown for 

 a distance of 4 mm. ; the whole truss was dead before the end of April 

 and on May 5 a canker extended from the base of the spur to three- 

 quarters of the distance round the branch. 



(6) Flower of a spur 2-5 cm. long inoculated April 8; another truss 

 of flowers borne on the same spur was untreated. The inoculated 

 truss was dead by April 26 and the disease had reached the base of the 

 other truss on that spur so that this also wilted. 



(c) and (d) Flowers inoculated April 8. Both trusses were dead 

 on April 26. 



The trusses thus treated were all borne on a branch bearing in all 

 six trusses of flowers, four of which were infected directly as shown ; 

 the fifth became infected indirectly from truss (6), and the sixth also 

 wilted early in May as the result of a canker produced at the base of 

 the inoculated truss {d) at the next node below. The cankers developing 

 round the bases of the infected spurs later became confluent and by 

 June 8 that portion of the branch bearins the trusses was cankered for 



