18 MISCELLANEOUS EESULTS OF WORK OF BUREAU IX. 



less number approached the poisoned baits than those not poisoned. 

 A series of tests with repellents showed that barriers of powder — 

 whether pyretlirum, arsenic, London purple, or Paris green — proved 

 obnoxious, the sowbugs quickly turning away to avoid the danger, and 

 showing, by the frantic waving of the antennae, that they had a per- 

 ception of something wrong. London purple seemed the least repel- 

 lent and yet practicalh^ as effective as any of the others. Sowbugs 

 placed in a jar with a biscuit rolled in arsenic became frantic and 

 died in a few minutes, as did others placed in jars with earth mixed 

 with either London pur]:)le or Paris green. 



After sprinkling Paris green under boards which had been favorite 

 haunts of the sowbugs, no more live specimens could be found, 

 although each day several dead ones were discovered. In April, 

 when the sowbugs were doing considerable damage to the cotton, a 

 mixture of Paris green and lime was dusted on and around the 

 sprouts with the result that under the poisoned j^lants great numbers 

 of dead sowbugs were found. No dead could be found around the 

 unpoisoned plants. The dusting was harsh treatment for the plants, 

 being in many cases fatal. It is, however, as proved by other tests, 

 unnecessary to dust the plants. The poison will be picked up by the 

 sowbugs in foraging over the ground. 



Under a roll of wire matting in his back yard the writer found the 

 soAvbugs so abundant that they crawled over each other in their haste 

 to get away. Having very little poison on hand, he sprinkled what 

 he had of Paris green, London purple, and arsenic over the ground in 

 an area of about 1 square yard and rolled back the matting. Next 

 morning he found 21 sowbugs alive and over 800 dead. Those alive 

 died in a few days, apparently from the efl'ects of the poisoning. 

 The poison washed from these dead sowl)ugs and used to saturate the 

 soil in jars in scA^eral experiments proved fatal to all sowbugs placed 

 in the jars. 



Kerosene emulsion as a contact spray was fatal. In spraying a 

 water barrel with kerosene the writer generally sprayed the ground 

 around it also, Avith the result that the sowbugs were ahvays killed. 



These experiments and tests Avere supplemented by numerous ob- 

 serA^ations of actual conditions from Avhich also data may be deriA^ed 

 regarding means of control. 



Concerning the plant-feeding habits, definite proofs Avere obtained 

 as folloAvs : 



May 25, at 7.30 a. m., soAvbugs Avere noted at various distances 

 aboA'e the ground feeding on the foliage of Aveeds and honeysuckle. 

 On June 30, at 7 a. m., three soAvbugs Avere discovered feeding on 

 weeds, and one at 8 feet above the ground feeding on a honey- 

 suckle leaf. Nine others Avere found on the honeysuckle vine at var- 

 ious heights up to 3 feet; also tAvo on grass blades and seA^en on the 



