426 Journal of Agricultural Research voi. vi.no. n 



In these five direct tests stomatal infection did not occur, thus con- 

 firming the observational data during the past two seasons and Dr. 

 Smith's earlier observations and experiments. In hundreds of field and 

 greenhouse observations the stems and leaves of wilted and healthy plants 

 were closely intertwined, exposing in many cases uninjured healthy parts 

 to direct contact with cut and broken infected surfaces. Even here the 

 disease was in no case transmitted. 



DISCUSSION OF OBSERVATIONS 



The field observations of the senior writer during the last two seasons, 

 covering the territory from the District of Columbia to eastern L-ong 

 Island, northward to the Canadian Provinces of Quebec and Ontario, 

 and westward to Michigan, Wisconsin, and Indiana, have abundantly 

 confirmed the experimental evidence outlined above that the striped 

 cucumber beetle and probably also the twelve-spotted cucumber beetle 

 are the principal if not the only carriers of bacterial wilt of cucurbits. 

 It has been suggested that the larvae of cucumber beetles may also serve 

 as a means of dissemination, but from their habits it would appear that 

 the only possible way in which they could bring about infection is by 

 carrying the organism from the soil into their burrows in the cucumber 

 stems. This appears highly improbable. However, the data at hand 

 do not warrant any definite statement. 



Mechanical injuries, such as those resulting from storms, cultivation, 

 etc., and injuries from flea beetles, aphids, and squash bugs have been 

 closely watched in the experimental fields and cages described elsewhere, 

 but no evidence has been obtained of any relation to bacterial wilt. 



WILT CONTROL 

 The problem of control therefore resolves itself into (i) the finding 

 or developing of cucurbit varieties resistant to bacterial wilt, (2) spray- 

 ing the plants with a bactericide, or (3) eliminating the beetles through 

 poisons or repellants. 



VARIETY TESTS 



Early in the spring of 191 5 a preliminary test was made with eight 

 varieties of cucumber planted in pots in one of the department green- 

 houses. Several plants of each variety were inoculated by needle punc- 

 tures in the leaves from 6-day-old agar-slant cultures of a single strain 

 of B. iracheiphilus. All the inoculated plants contracted the disease 

 and no difference in rapidity of wilting appeared — that is, individual 

 plants of the same variety showed as great differences in rate of wilting 

 as appeared among the different varieties. 



In the variety-test block previouslv mentioned (East Marion, Long 

 Island, N. Y.) 32 varieties of cucumber, 39 varieties of cantaloupe, and 

 25 varieties of squash were planted on June 10, 191 5. From 8 to 20 or 

 more hills were given to each variety, 12 being the usual number. Most 

 of the cucumber and squash varieties gave fair to good stands, but the 

 cantaloupes were planted in an exceedingly light sandy soil infested 

 with witch grass, and in consequence of this the seed either did not come 



