BACTERIAL WILT OF CUCURBITS. 25 



arose, which at first somewhat resembled Bacillus tracheiphilus. Later, on beef-agar 

 slants, this type gave a dense, opaque, somewhat yellowish and rapidly growing 

 colony. A second type (B) was constantly found in the plates poured from one set of 

 dilutions. This isolation (J5) more closely resembled the wilt organism, and young 

 colonies on beef agar were at first scarcely to be distinguished. Even the rate of 

 growth was about the same as for Bacillus tracheiphilus. However, old colonies are 

 somewhat more opaque and also have a very slight yellowish cast not present in the 

 wilt organism. The organism also produces more alkali than B. tracheiphilus. Inocu- 

 lation tests on cucumber plants have given negative results with both these types of 

 bacteria. Occasional colonies of other bacterial and fungus types came up on the 

 plates, but on account of their rarity were considered as contaminations. The wild 

 organism was not obtained. 



Portions of intestines left from each beetle dissected were retained and in each 

 case were used in inoculating a single young cucumber plant by needle puncture. 

 One plant only out of the ten wilted, but isolation was not attempted. There were 

 at the time a large number of other cucumber plants in the greenhouse, all of which 

 were free from wilt, and there is little doubt that the plant in question had become in- 

 fected with the true bacterial wilt. 



It was thought worth while to test the types of intestinal bacteria in mixet 

 culture with authentic isolations of Bacillus tracheiphilus for possible antagonism. 

 After three weeks in mixed cultures (beef agar plus 1 per cent cane sugar) young 

 cucumber plants were inoculated by needle punctures in the leaves; and other plants 

 were inoculated as checks with intestinal types A and B alone and with the wilt organ- 

 ism alone. Of ten plants inoculated with cultures of A plus wilt, and of the two plants 

 with A and B alone, none became infected; but of four plants inoculated with B plus 

 wilt all contracted wilt, showing the typical stringy ooze from the cut stems and the 

 vessels filled with bacteria. All plants inoculated with B. tracheiphilus alone showed 

 wilt. The plants were under observation one month. 



In another experiment using four plants with type B alone, four plants with the 

 wilt organism alone, eight plants with A plus wilt, and four plants with B plus wilt, 

 all from 2-weeks-old cultures, the results were identical — growth of B. tracheiphilus 

 and infection in the presence of B; none in the presence of A. 



The wilt organism itself was not isolated on agar-poured plates 

 made from beetle intestines, but evidence was adduced that the 

 organism was present in the intestines of one of the ten beetles used 

 and that in the presence of cane sugar the intestinal flora of at least 

 one beetle was not antagonistic to the wilt organism, but that in the 

 majority of the cases the intestinal flora (type A) was under these 

 conditions antagonistic to the wilt organism. 



These results, while not sufficient to warrant far-reaching con- 

 clusions, suggest the possibility that a careful study of the bacterial 

 flora of cucumber-beetle intestines will afford an explanation of the 

 fact that only a small proportion of wilt-fed beetles are wilt carriers. 



VIRULENCE TESTS. 



During the interval since August, 1914, several hundred isolations 

 of Bacillus tracheiphilus have been made from cucumber, cantaloupe, 

 winter squash, crookneck summer squash, and cymling, and from an 

 area extending from Canada to southern Georgia and Long Island to 

 Iowa. Of these isolations from various hosts and localities, over one 



