Wormald.-- Brown Rot ’ Diseases of Fruit Trees. 371 
branch the lesion tends to become healed quite early (Fig. 16), and no 
more pustules develop on the cankered surface of the branch itself subse- 
quent to the year following infection. When, however, the canker girdles 
and kills the branch the lesion cannot be healed and the fungus may 
persist in the tissues of the canker for several years. In the spring of 1917 
large cankers bearing numerous pustules were removed from affected trees ; 
some of these were placed in a pot on the ground, others were suspended 
at about 3 feet above the ground level. In January, 1919, those cankers 
which had been suspended still bore numerous powdery pustules (see 
Fig. 12) ; of those on the ground most had lost their bark and no pustules 
were present, while the rest bore a few pustules at the upper end. Such 
cankers, therefore, may serve as sources of infection for at least three 
years. 
Another feature, not previously recorded, in connexion with this 
disease is the fact that when a leading branch is girdled by a canker, 
which, as already shown, 1 occurs about the end of June when the tree is 
growing vigorously, the upward flow of sap is suddenly checked and 
diverted into the buds immediately below the canker, with the result that 
these buds are stimulated to grow out into weak vegetative shoots instead 
of developing into fruiting spurs (Fig. 15). A similar result has been 
described and figured as occurring on plum trees affected with the ‘ Wither 
Tip ’ disease. 2 On pear trees I have met with Monilia cinerea on one 
occasion only, when it was found on a young pear about 1 cm. in length, 
while M.fructigena often occurs, causing a brown rot, on the ripening fruit. 
Pyrus japonica is subject to a Blossom Wilt similar in general appear- 
ance to that of the apple tree. I have isolated the fungus in one case, and 
it proved to be similar physiologically and biologically to the form of 
M . cinerea obtained from plums and cherries. 
IV. Methods adopted in isolating Pure Strains. 
The strains of Monilia used in the experiments recorded in this paper 
were all c pure ’, each being derived from a single conidium ; the conidia 
were obtained for this purpose either directly, from naturally infected 
specimens bearing fructifications, or indirectly from cultures prepared with 
the object of inducing the development of conidia in those cases where 
mycelium only was present in infected material. 
A method of isolating pure strains of Monilia from infected tissues 
containing mycelium has been described in previous papers ( 1917 - 18 ), and 
brief references have there been made to a method of obtaining strains 
from single conidia. The latter given in detail is as follows : Two watch- 
1 Vide Ann. Appl. Biol., vol. iii, p. 167. 
2 Vide ibid., vol. v, Plate VIII, 'Fig. 3. 
