20 



eye — i. e., the disease seemed to have died out in the parts above 

 ground. In the other two there were distinct symptoms in the bulb. 

 In one bulb several scales had yellow bundles, and the plateau was 

 also diseased in the upper j^art ; in the other bulb the disease was 

 restricted to two bundles in the upper part of one scale. 



SERIES 4 (ONIONS). 



Three onion plants {Allium cepa) were selected for this series, 

 which was begun Januaiy 29, using the same culture medium and 

 method of inoculation as in series 3. Each jslant was copiously 

 inoculated in the aj)ex, middle, and extreme base of several leaves. 



Result. — The young and tender leaves were killed outright within 

 a few days of the inoculation, with no distinct symptoms of parasit- 

 ism. The older leaves developed no symptoms whatever, or only 

 such as were due to the slow growth of the parasite in the immediate 

 vicinity of the point of inoculation — i. e., the symptoms were entirely 

 unlike those obtained by Heinz with his Bacillus hyacinthi-septicus. 

 In case of half a dozen or more leaves the germ was able to hold its 

 own in the inoculated tissues and finally to make a bright yellow 

 growth in the parenchyma in the vicinity of the punctures. It never 

 extended very far, however, and did not kill the parts in which it 

 grew — at least not until after many weeks. PI. I, fig. 2, shows the 

 appearance of an onion leaf in which the germ has made a slow 

 growth. 



On June 22 these plants were knocked out of the pots and their 

 bulbs examined. Each plant had a good top at this time and was in 

 fruit. One had seven bulbs from a common root, another four, and 

 the third three. xVU of these bulbs were sound. None showed any 

 trace of yellow bundles. 



SERIES 5 (HYACINTHS). 



The fifth series of inoculations was made February 7, in the same 

 way as the two preceding. These plants were inoculated from an 

 alkaline beef broth culture (No. 1, Jan. 20), about 0.5 to 0.7 cc. of 

 cloudy broth being used on each plant. Nine vigorous plants in full 

 bloom Avere selected for this experiment, the variety being a single- 

 flowered, pale-blue sort known as Czar Peter. Two were inoculated 

 in the scape just under the inflorescence (0.3 cc. each, several punc- 

 tures) and the remainder were inoculated in the apical portion of the 

 leaves, three to seven leaves on each plant being selected for this 

 purpose (generally three leaves).. Twenty-three plants of the same 

 variety and growing in the same box were held as checks. 



Result. — Ten of the check jDlants were attacked by a soft white rot 

 between Februaiy 7 and June 14, The bulbs of three of these were 

 only softened a little in places when dug out and these bulbs showed 

 no trace of yellow bundles. The other seven were destroyed by the 



