Murphy — Bionomics of the Conidia of Phytophthora infestans. 443 



experiment indicated that by so doing the spores remained viable and capable 

 of infecting potatoes for several days. 



On September 15th the following experiment was begnn. Finely sifted 

 potting loam, in a comparatively dry condition, was dusted on the lower 

 surface of potato leaves which carried a strong growth of conidiophores of 

 Phytoplithora infestans, and was then brushed off again with a fine brush. 

 In this way a quantity of soil, weighing about 400 grams, was so copiously 

 contaminated that it was easy to find conidia in any fraction when examined 

 microscopically. The soil so prepared was divided into four parts, each of 

 which was placed in the form of a little mound on the surface of similar 

 loam which filled a flower pot. Four similar pots were prepared in which 

 were placed like quantities of sifted potting loam which had not been 

 contaminated. All the pots were placed on a roof exposed to the weather 

 but shaded from the sun for part of the day. The length of time the conidia 

 survived was determined by inoculating a little of the soil from the mounds 

 into shallow wounds in healthy tubers.^ 



Using this method, it was found that the fungus remained capable of 

 infecting tubers for twenty-one days, but not for twenty-eight days. The first 

 inoculations on the unwounded surface were failures ; thenceforth they were 

 made in wounds, with the result that on the loth day, 8 out of 8 inocula- 

 tions were successful ; on the 19th day, 4 out of 8 ; on the 20th day, 4 out of 

 6 ; on the 21st day, 3 out of 6 ; on the 28th day, out of 8. Throughout 

 the experiment no disease developed in any of the thii-ty-two wounds in 

 which was placed soil similarly exposed but not contaminated with conidia. 



This experiment was conducted during the driest portion of the year, 

 from September 15th to October 6th, 1920. Notwithstanding their exposed 

 position the conidia survived this. Torrential rain occurred on October 2nd 



' Examinations were first made microscopically, but it was found that searching for 

 viable conidia in soil, and then testing their germination, was a time-consuming opera- 

 tion. Reliance was, therefore, placed on infection experiments on potato tubers with 

 adequate controls. At first whole tubers were used, the inoculum being introduced 

 into wounds, while similar potatoes from the same lot were wounded but not inoculated. 

 The bulk of the work, however, was carried out on potato slices, and a portion from 

 every tuber used was left untreated. In all the critical experiments an "adjacent 

 surface control " was allowed to every inoculated slice, that is, the two surfaces which 

 were in contact before a tuber was cut were used as a check on each other, one being 

 inoculated and the other not. It may be stated that of the 232 control pieces used 

 throughout all the experiments to be described, only four failed to remain sound, and 

 none developed blight. In no single case did non-contaminated soil produce blight, and 

 it was always used for comparison with contaminated soil. In all the work, with the 

 exception of some of the preliminary experiments (in which the typical appearance of 

 blight rot was deemed sufficient), infection was not admitted until the parasite was 

 recovered from the infected material. 



