1-2 



ieet wide. One-half was kept as a check and, beginning when the first 

 squares were formed, Paris green was applied by means of the sack- 

 shaking method on the following dates: 



Pounds 

 I)er acre. 



June 10 1 



June 16 1| 



June 2-4 : IJ 



July 5 11^ 



July 15 1\ 



August 1 1] 



From Jul}^ 13 to August 3 examinations were made to ascertain the 

 relative number of squares infested on the poisoned plat and its check. 

 Blooming ceased before the first examination was made, and at practi- 

 cally the same time on both plats, so the table does not show the rela- 

 tive infestation while blooming was in progress, but it does show most 

 conclusively that the poison did not kill enough weevils to allow the 

 treated plat to keep on ])looming and producing bolls after the untreated 

 check had stopped blooming. 



T.^ULE II. — E.ramiiiadoiis of Paris-ijrcoi experiment, San Antonio, Tex. 



An average of all examinations on Paris-green plat shows 89.94 

 per cent infested and on the check 89.53 per cent. The fact that the 

 percentages are almost identical shows that the poison had practically 

 no effect. 



The poisoned area yielded 190 pounds and the check 93 pounds of 

 seed cotton, an increase on the poisoned plat of 97 pounds, which would 

 be a gain of 113 pounds per acre. The cost of poisoning was about 

 $1.80 per acre, so that, making no allowances for the circumstances 

 noted in the following paragraph, the net gain was onl}^ about $2.15 

 per acre, reckoning seed cotton at 3^ cents per pound. 



Running diagonally across the poisoned plat was an old road, the 

 presence of which was not known when the experiment was planned. 

 This road touches only one corner of the check plat. Along the loca- 

 tion of this old road the cotton grew very rankly and produced much 

 more abundantly than did the adjacent portions of the field. Making 

 an allowance for the increased production on the Paris-green treated 



