10 N. J. Agricultural Experiment Stations, Bulletin 355 



killing- strength of gas in greenhouses under which a trench to 

 acccMiimodate pipes for heating is present. In this fumigation 

 a hole was cut through the concrete floor of the greenhouse and 

 one of the tubes was placed with its opening 1I/2 feet below the 

 floor, so that it projected down into the trench. The result of 

 the fumigation is seen on the chart. The gas apparently dif- 

 fused downward and entered the trench, where it attained the 

 highest concentration of any point at the end of the fumigation. 

 This finding would suggest that failure to secure killing with a 



I Z 3 ^ 5 6 7 



Q 9 10 II /£ 13 14 15 le 17 13 19 20 2/ 22 

 MINUTES 



Fig. 3. Graph op Fumigation No. 3 



A, outside bench; B, middle bench; C, center bench, D, mid- 

 dle bench; E, outside bench; F, outside collecting tube; G, middle 

 collecting tube; H, center collecting tube; I, comb collecting tube. 



charge adecpiate to get kill in a house of a given cubic capacity 

 may be due to a loss of gas downward into a trench beneath 

 the house. In this case the house was reglazed and tightened 

 before the fumigation so that it was in the best of condition as 

 regards leakage, but nevertheless there was a great loss of gas 

 through the trench beneath. 



Figures 13 and 14 showing the results of tests with a 3-ounce 

 and a 6-ounce charge, respectively, show very well the course 

 of the fumigation. The great uprush of heated gas and the sub- 

 sequent diminution of concentration as the fumigation proceeds 

 is well shown. Likewise, a peculiarity of the house in which 



