40 BULLETIN 395, U. S. DEPAETMEXT OF AGRICULTURE. 
manner suggesting infection from rain-beaten or wind-blown spores; 
and (4) that even under these severe conditions no infection occurred 
upon the protected surfaces of the fruits, doubtless due to the fact 
that these areas are rarely, if ever, thoroughly wet during the more 
important periods of fruit infection, and consequently both the trans- 
fer of spores to such areas and the subsequent development of the 
fungus are practically precluded. 
CONCLUSIONS. 
From the evidence presented, it appears to be conclusive that a 
large majority of fruit infections result from water-borne spores from 
lesions on the subtending twigs. Considerable dissemination must 
obviously be accomplished by meteoric water which drips or spatters 
from spore-bearing areas. Wind dissemination evidently occurs to a 
limited degree, especially in cases where spores are detached from the 
parent fungus by chance mechanical agencies, as by the rubbing 
together of organs of the host plant or by the impact of wind-blown 
sand particles. It is altogether probable that other general agencies 
of spore dissemination, such as insects and birds, may play some part 
in spreading the disease, but no evidence of this has been observed. 
If it occurs, it is certainly of very minor importance, except, possibly, 
as a means of transferring spores over relatively long distances. 
METHOD OF INFECTION. 
The exact method by which the fungus penetrates the cuticle of 
the various affected organs of the host has not been conclusively 
determined. TVhile the solution of this problem is necessary to a full 
knowledge of pathogenesis, the attendant difficulties are such that this 
would have required the expenditure of more time than would be 
justified by its importance to this paper. Certain observations and 
experiments in this connection, however, seem to be worthy of record. 
The abundance of naturally and artificially induced infection shows 
conclusively that the penetration of the fungus is independent of 
woimds. Consequently, it must normally occur through natural 
openings, directly through the cuticle, or in both of these ways. Since 
the problem is presented in its simplest aspect hi leaf infection, 
certain studies of leaf penetration were made, as follows: 
Experiment 1. — On July 15, 1915, a potted 2-year-old Early Crawford tree bearing 
abundant foliage in various stages of development was inoculated by spraying with 
a water suspension of spores from a 15-day-old Lima-bean agar culture of a twig strain 
of the fungus. The plant was kept in a moist compartment and sprayed daily with 
sterile distilled water. Marked inoculated leaves in various stages of development 
were collected daily and placed in stoppered test tubes in a solution prepared by 
mixing glacial acetic acid and 95 per cent ethyl alcohol in equal parts. After the 
leaves were thoroughly bleached, their under (dorsal) surfaces were carefully studied 
microscopically, previous inoculation experiments having shown that the upper 
