2i6 Journal of Agricultural Research voi. xv. no. 3 



of one of the spots on one of the cotyledons showed bacteria in the intercellular spaces. 

 Two of the seedlings were preserved as herbarium specimens. On March 27 another 

 infected seedling was noted in another flat. The organism was isolated and identified 

 by inoculation as before. 



Experiment of March 27, 1917.' — Fourteen flats of sand and four of heavily com- 

 posted garden soil were steamed for one hour at 7 pounds' pressure. All but two were 

 planted with the seed from Ohio. These two planted with seed treated in 1916 with 

 mercuric chlorid and untreated seed from the 1916 supply, respectively. Precau- 

 tionary measures taken as before. On April 4 a typically infected seedling was noted 

 in one of the sand flats planted with the Ohio seed and on April 7 a well-advanced 

 stage of the disease was discovered on a seedling in another sand flat of the Ohio seed. 

 There was no doubt as to the cause of the lesions from the characteristic signs — viz, 

 water-soaked tissue and white exudate residue. Platings from each of these seedlings 

 gave an abundance of the typical colonies. 



The results of these experiments and the fact that in Dr. Gardner's 

 later tests of the Ohio seed in sterile damp chambers one seedling in each 

 of two dapip chambers developed the typical signs of the disease prove 

 that the angular-leafspot organisms may live for at least seven months 

 on the seed. There seems no reason to doubt but that they can survive 

 for two months longer and infect the seedlings as field observations have 

 indicated. 



The use of seed as badly contaminated as the lot from Ohio was found 

 to be, would have resulted in the early development of angular-leafspot 

 in as large a proportion of the fields as occurred in 191 6 in Wisconsin. 

 From the Ohio lot approximately 3,500 seeds were planted with the pre- 

 cautions described. Seven, or a proportion of i to 500, of the resulting 

 seedlings developed angular-leafspot. With this proportion or 0.2 per 

 cent and the use of 2 pou'nds of seed per acre, as is usually practiced, there 

 would be about 72 plants infected from seed-borne organisms to every 

 acre of cucumbers. 



As to how the organisms are protected on the seed so as to withstand 

 the long period of desiccation there is no conclusive evidence. It seems 

 most likely to the writer, however, that they get in at the micropylar end 

 of the seed, and so are protected within the seed coat. The fact that the 

 iijfections of the seedlings nearly always occur on the edge of the coty- 

 ledons near the point of attachment to the stem — the part of the coty- 

 ledons which is at the micropylar end — indicates that the bacteria are 

 probably harbored beneath the seed coat (PI. 16, A). It might be 

 argued that, since on germination the attached ends of the cotyledons 

 are the first to emerge, the portion which becomes infected is the first 

 part which is exposed to organisms on the surface of the seed. This ex- 

 planation, however, seems less probable to the writer- than that the or- 

 ganisms are sheltered inside the micropyle. At any rate subsequent 

 work by Gardner and Gilbert {8) has shown that the bacteria are so lo- 

 cated that they can be killed by chemical treatment of the seed. 



1 Performed in cooperation with Dr. M. W. Gardner. 



