10 SEED STERILIZATION AND ITS EFFECT UPON INOCULATION. 



The rapid mortality of bacteria on the sterilized seeds shown in 

 this series indicates that some influence Avas acting to weaken their 

 normal vitality and that this varied with the vigor of the culture." 

 It further appears that the absorption of the liquid culture by the 

 soil and by the filter paj^er used for drying removed the bacteria 

 largelj^ from the seed coats, especially in the case of peas and alfalfa. 

 Seed dried in open petri dishes gave nearl}^ as good returns as seed 

 dried in the desiccator. These tests were repeated with practically 

 the same results. A greater viability of legume bacteria on washed 

 but unsterilized seed is reported in some similar tests by Edwards 

 and Barlow.'' 



To test the possible removal of external contaminations b}' wash- 

 ing without disinfectants, several lots of leguminous seed (fourteen 

 in all, of different kinds) were allowed to soak in steril;e water for 

 two hours, then rinsed twice, and the third water allow^ed to stand 

 on the seed for one hour. Platings were made from the first and 

 third waters, inoculating one loop into agar. Six of the fourteen 

 lots gave sterile plates from the first water, doubtless due to the 

 very small quantit}^ of material used for inoculation. From the 

 third water sterile plates resulted in every case but one, which had 

 a single mold colony. 



The seeds were allowed to soak two days in the third wash water 

 to start any remaining organisms into growth. They were then 

 placed in sterile petri dishes and germinated by pouring over them 

 sterile melted agar. As was expected, nearly all of these tests gave 

 plates somewhat contaminated — much less, however, than seed germi- 

 nated similarly but without previous rinsing. 



A third series of tests was then made, in which the vacuum method 

 of washing was used, each lot of seed being rinsed five times. These 

 seeds when germinated in agar showed less contamination than those 

 in any of the preceding series, but a few organisms persisted in nearly 

 every case. 



As some of the plates showed colonies somewhat resembling the 

 legume organism (Pseudomonas radicicola) , a further test was made 

 in which seed previously treated with pure cultures was used to 

 make certain that Pseudomonas radieicoJa was originally present. 

 One half of the treated seed was washed by the vacuum method, 

 receiving, however, five rinsings instead of four. The other half 

 was simply moistened and germinated, both lots of seed being germi- 

 nated by pouring over them in sterile petri dishes melted nitrogen- 



" Similar results showing increased resistance to killing out when legume 

 bacteria were dried in a desiccator were reported by Kellerman and Beckwith 

 (Science, n. s., vol. 215, 1900. pp. 471-472). 



& Bulletin 169, Ontario Agricultural College, 1909, p. 15. 

 [Cir. 67] 



