MASS. EXPERIMENT STATION BULLETIN 207. 



Application. 



The trees used were the apple, cherrj', peach, pear, plum and elm. The 

 materials were applied in the same way in all cases, being thoroughly mixed 

 Mith the proper amount of water just before using. With the acid pastes, 

 3 pounds in 50 gallons of water, and with the powder, 1^ pounds, were 

 used, the powder containing approximately t\vice as much arsenic pen- 

 toxide as the pastes. As the neutral arsenate contained much less pen- 

 toxide than the acid pastes, 5 pounds 7.6 ounces of it were mixed ^\^th 50 

 gallons of water to provide an amount of arsenic pentoxide in the spray 

 equal to that present in the others. Practically an equal amount of 

 poison was therefore applied in every case. 



It has been suggested that injury might be caused by the poison entering 

 the leaf through the stomata. As these are usually more numerous on the 

 lower than on the upper surface, branches were held by the hand in such 

 a position that the spray would reach only one surface of the leaf. Par- 

 allel tests for both surfaces were made, one test immediately following the 

 other and on the same tree. The main lines of investigation, though, were 

 with reference to variations of temperature and humidity and of light. 

 Two series were made, one in bright, clear weather, and the other on 

 cloudy days. 



The temperature and humidity were obtained from a Hygrodeik manu- 

 factured by Andrew J. Lloyd & Co. of Boston, giving both the temperature 

 and relative humidit}'. These were taken at the tree immediately before 

 applying each spray. The attempt was made to spray each surface of the 

 leaves, both in clear and cloudy weather, for at least every 5° interval be- 

 tween 65° and 95° of temperature, and between 50° and 90° of humidity. 

 To obtain all these combinations, however, proved difficult, and some of 

 them were not obtained until several years had elapsed, though fairly 

 complete series were finally secured. 



Application of the sprays was begun in June, continued during July, 

 and a few sprays were put on the trees early in August. The tests were 

 begun in 1912 and ended in 1920. After the spray had been applied, its 

 effect was observed about twice a week for at least two weeks, so that any 

 injury appearing late might not be overlooked. 



Adequacy of Experimental Methods. 



Three possible sources of error, at least, may have affected this experi- 

 ment. First, there is the difficulty of a uniform estimation of the amount 

 of injury found. As a check upon this we have the very uniform agree- 

 ment in observations made at identical and nearlj^ identical temperatures 

 and humidities, not only in the same, but also in difYerent years. The 

 personal equation was reduced as much as possible by having the obser- 

 vations all made by one person. Then, after all, the main dividing line 

 was between injury and no injur}', determination of the exact degree of 

 injury being of less importance. 



