EXPERIMENT STATION BULLETINS. 485 



leased from the collars. Proof was soon obtained that this would not 

 do; the insects licked the joints of their legs and other rough parts of 

 the body so thoroughly that the contents of the crop became decidedly 

 stained. Jn other words, the insects did their cleaning so carefully that 

 they might be eating enough of the borax powder to kill. The only safe 

 method, therefore, was to leave the collars on the animals until they 

 were no longer able to lick various parts of the body. That the death 

 of an insect so treated was due primarily to the influence of borax 

 powder brought in contact with the bod}', and not to injury from the 

 collar (or oilierwise) was shown by the checks carried out with un- 

 treated insects in collars, and by the absence of any stain in the digestive 

 tract. 



The evidence gained from the experiments carried out in this connec- 

 tion, therefore, nmy be summed up in the following statements : 



When cockroaches run through powdered borax, sodium fluoride, or 

 borax and pyrethrum mixed, the dry i)owder sticks to the legs, body, 

 and often to the antennae. Some of the powder is moistened or dis- 

 solved in the exudati(m about the bases of the legs and on the thinner 

 portions of the outer integument. This seems to cause some irritation 

 and uneasiness; the insect soon begins to clean the moistened iJOAvder 

 from the body by licking it. In doing this, enough of the jjoison may be 

 brought into the moutb, and swallowed, to kill after a period varying 

 from Uve hours to ten dajs. JJorax acts much more slowly but kills 

 with no less certainty than sodium fluoride. It is possible for both 

 powdered borax and sodium fluoride to kill roaches by contact (without 

 any of the poison's having been swallowed), but in actual practice it is 

 hardly possible that any roach ever gets its body dusted with one of 

 these powders without also licking and swallowing some of the powdered 

 poison. Both powders are more rapid in their action as stomach poisons 

 than they are as contact insecticides. 



Tables IV and V contain representative results illustrating the in- 

 fluence of borax, sodium fluoride, pyrethrum and white hellebore on the 

 reductases and catalases, respectively, of insect extract. Borax acted 

 quite injuriously on the activity of the reductase, and very little on that 

 of the catalase. The same was true of the hellebore and pyrethrum, 

 except that the pyrethrum seemed to interfere a little more with the 

 activity of the catalase than did either the borax or the hellebore. 

 Sodium fluoride was most detrimental to the catalase. AVhen added 

 to perfectly fresh extract, the. sodium fluoride caused a slight drop in 

 reductase activity; after that however, such a treated extract retained 

 its reductase activity very much better than did untreated extract. 

 Both sodium fluoride and borax interfered with the oxidase activity 

 slightly, if one may regard the alcoholic guaiac test as an indicator in 

 this case. 



An attempt was made to determine the influence of borax and sodium 

 fluoride upon the respiratory ratio of cockroaches dying under the in- 

 fluence of these poisons, but the normal respiratory ratio, itself, varied 

 so much for the roaches that no decided information could be gained. 

 The intestinal tracts of the specimens used were apt to be gaseous, and 

 the contents of the intestines were usually found to be well infested 

 with infusoria, etc. This, it was believed, might accoimt for the normally 

 irregular respiratory ratio. 



