juneio, i9i8 Experiments with Coccomyces spp. from Stcnie Fruits 565 



pennsylvanica was very limited, and it is possible that further trials may 

 yield positive results. The results from the strain from P. avium paral- 

 leled those from the P. cerasus strains, with the exception that no infec- 

 tion was induced on plums. It should be remembered, however, that 

 only one strain from P. avium was available, and its pathogenicity 

 appeared to have been diminished in culture. 



The strains from P. pennsylvanica infected P. avium, P. mdhaleh, and 

 P. pennsylvanica abundantly, and P. cerasus, P. munsoniana, and P. 

 cerasifera with difficulty. They differed in behavior from those from 

 P. cerasus chiefly in their abundant infection of P. pennsylvanica and their 

 sparse infection of P. cerasus, on which but i trial of 10 gave positive 

 results. It is important to note, however, that, while only one cross 

 from P. pennsylvanica to P. cerasus was effected, and the reverse cross 

 has not yet been made, crosses between P. pennsylvanica and P. mahaleb, 

 on the one hand, and P. cerasus and P. mahaleb, on the other, were 

 readily effected. Furthermore, P. pennsylvanica readily infected P. 

 avium, which easily cross-infects with P. cerasus. It appears, therefore, 

 that while the results from strains from P. pennsylvanica do not accord 

 perfectly with those from strains secured from P. cerasus, P. avium, and 

 P. mahaleb, the agreement is so close that, from the standpoint of host 

 relationships, the strains from these four species may tentatively be 

 considered as a single group in which varying degrees of specialization 

 have been developed. It remains, however, for further experiments to 

 define the extent and the constancy of this specialization. Attention is 

 called to the fact that, while the crosses just reported suggest the possi- 

 bility that strains from P. pennsylvanica may pass readily to P. cerasus 

 by way of P. mahaleb or P. avium, no experiments were made in which a 

 given strain was carried over in this manner. Such experiments are 

 projected. Further discussion of these problems of specialization and 

 a consideration of the specific relationships of the various strains of 

 Coccomyces used in these experiments are reserved for a later paper. 



The results from strains taken from P. domestica, P. virginiana, and 

 P. serotina are so different that, from the standpoint of host relationships, 

 no further grouping of strains seems feasible. Strains from P. domestica 

 infected P. domestica and P. americana abundantly, and P. salicina, 

 P. munsoniana, A. persica, P. besseyi, and P. mahaleb with difficulty. 

 Those from P. virginiana infected P. virginiana and P. padus easily and 

 abundantly, and P. mahaleb consistently (except one trial), but uniformly 

 less abundantly. They infected P. domestica and P. insititia with diffi- 

 culty. The strains from P. serotina showed a high degree of speciaUza- 

 tion. They infected P. serotina uniformly and abundantly, P. mahaleb 

 fairly consistently but sparsely, and P. insititia with difficulty. 



A condensed summary of the results of all the crosses tried appears in 

 Table X. Of all the experimental plants, P. m,ahaleb showed the widest 

 range of susceptibility, being infected by inocula from all the host sources 



