n. T. W lUUKKR AND (). (J ROVE 01 



been taken to ascertain that the flowers showed no sign of such damage 

 before being selected for the experiments under cover. On the other 

 hand many cases occur under outdoor conditions in which it is difficult 

 to decide whether the damaged blossom has been affected by the organ- 

 ism or by frost, the type of damage appearing to the naked eye very 

 similar in either case. Again in other instances there is no question of 

 the damage being due to frost. In this connection it may be noted that 

 the blackening or browning of the pistils in unopened or partially opened 

 buds can be caused by frost even when the other parts of the flower are 

 quite unaffected. Striking cases of this kind have been observed this 

 spring on the Myrobella plum. The bacillus having been isolated from 

 frosted blossoms, from flowers with blackened pistils which may or may 

 not have been caused by frost, and from undamaged fully expanded 

 flowers of the Myrobella, it is evident that for this plant at least the 

 organism is not always pathogenic. At the same time the tissues of 

 the discoloured pistils have been swarming with cells of the bacillus 

 in some instances. Further investigation is necessary before such points 

 can be satisfactorily cleared up. 



When the disease was at its height last year in the fruit plantations 

 at this institution one of its most striking features was the rapidity with 

 which it spread from flower to flower. Definite proof was forthcoming 

 that this was due mainly, if not entirely, to the agency of bees and other 

 insect visitors to the flowers. A number of bees were caught in sterilised 

 test-tubes, while they were actually working among trusses of pear 

 blossom. They were transferred to Petri dishes containing a layer of 

 sterile malt extract gelatine, and were allowed to walk over the surface 

 of the latter. In fifty per cent, of the cases examined it was found that 

 colonies of the bacterium with the typical characters to be described 

 later developed in the footprints of the bees after an interval of three or 

 four days. The course of the bees across the plate was most strikingly 

 mapped out by the fine of colonies. It is interesting to note that per- 

 fectly pure cultures were obtained in some cases in this way, no 

 other organism developing on the plates. 



There is no doubt therefore that the dissemination of the disease is 

 largely due to insect visitors to the flowers. Infection is carried by them 

 from diseased to healthy blossoms, which become inoculated either 

 through the stigmas or the points of the viscid receptacle with which 

 the feet of the insects come into contact. It will be seen later that 

 infectioQ can take place by merely superficial inoculation with the 

 organism in this manner. 



