648 Journal of Agricultural Research vca. x, no. 12 



elusion that these facts hold true in the field during any rainfall; and, 

 in view of my previous experiments showing the emanation of motile 

 bacteria from lesions into the surface water film, their dissemination is 

 the natural result. 



It was stated in the paper cited above ^ that — 



imless a great amount of this spread took place at one time, it would have been diffi- 

 cult to imderstand how the disease could progress so far in one direction if each spot 

 took 8 to 10 days to appear. 



And the second inoculation experiment in which bacteria were used 

 which had never been in culture failed to clear the question. The data 

 presented here, however, show that it was entirely possible for most of 

 the spread to have taken place at one time, though no records of wind 

 velocity during the two thunderstorms of June, 1916, are available. 



In the experiment in which the effect of the wind was studied all 

 the factors were within the limits of field conditions. A drop of 0.02 c. c. 

 in volume falling 16 feet upon a relatively thin film of water which was 

 splashed, as is proved by the acetic-acid distribution, upon a plate 3 

 feet above the floor during a wind of 10 miles an hour, splashed water in 

 abundance a distance of 8 feet (across two rows of cotton) in moderate 

 quantities as far as 12 feet (three rows) and in slight amounts to 16 feet. 

 To quote again from the former paper : ^ 



The possibilities of this chain of action during a driving rain are considerable if 

 one includes the distance bacteria may be carried from the original lesion, then 

 splashed up again and carried farther, and so on, until a dilution too great for infec- 

 tion is obtained. 



Local dissemination of such diseases as described by Rolfs, Pierce, 

 and others is well within the limits of probability in the presence of dew 

 and heavy fogs without v.dnd. Each drop (the drops from leaves are 

 larger than the average raindrops) falling 1 2 inches upon a film of water 

 will scatter splash drops over an area of 20 to 32 inches in diameter. The 

 rapidity with which local dissemination of such diseases takes place is 

 easily accounted for. 



One could well expand this discussion to include the applications of 

 these conclusions to other similar diseases, but, since no opportunity has 

 been had to study the conditions under which such dissemination must 

 take place, this phase is omitted for the present. However, one can 

 readily understand the possibilities if he takes into consideration the 

 increase of extreme distance of splash due to greater elevation of surface 

 of impact as would occur in pear, walnut, and Citrus trees and greater 

 wind velocity, as occurs in typical thunderstorms of this region and 

 especially during the hurricanes frequent in the southeastern Citrus 

 regions. 



1 Faulwetter, R. C. Op. cit., p. 467. * Idem, p. 470. 



