226 REPORT 4, UNITED STATES ENTOMOLOGICAL COMMISSION. 
brushes throw the poisoned liquid as a spray from the cups. I have 
not seeu the operation of this machine, but judge from a similar device, 
arrived at independently and experimented with, that it contains a 
principle which in a modified form will be of some value in poisoning 
cotton-insects. 
The following statement should be appended here. In speaking of 
the means of applying Paris green in water to the cotton crops, Dr. 
George E. Gillespie, of Natchitoches, La., reports: 
" The mode of applying it differs according to the means of the planter 
using the poison. Many persons make brooms of mayweed {Anthemu 
>cotula), which are handy and very convenient; others use the common 
garden watering-pots." — Bull. No. 3, p. 125. 
Rotated recesses. — These, in the ends or sides of rotated bodies, 
may be used to throw liquids and powders in a centrifugal manner. 
Where such devices whirl very rapidly it is difficult to feed them satis- 
factorily upon the periphery, and I found of most value a sheet-metal 
funnel with internal radiating septa. This is rotated rapidly while sup- 
plied inside with a dripping stream.' 
After devising and using simple apparatuses to test the spraying 
powers of rotary throwers of the various kinds noticed in this chapter, 
I concluded that another class of machines was superior, and hence 
stopped the study of such throwers to take up that of blowers, which 
follows. 
III. — BLOWERS OF POISON. 
Poisonous gases, vapors, liquids or powders, as insecticides, can be 
■applied by means of various blast-discharging or blowing devices, and 
•examples of apparatuses for such purposes will be noticed below. 
ROTARY BLOWERS OF POISON. 
[Plates XXVIT, XXVIII, XXIX.] 
In order to show the efficiency of rotary fans for blowing poison I 
designed several different styles, including simple machines to be worked 
hy hand, as well as a compound one for horse-power. Most of those 
made and tested proved successful. 
The power required to drive one of these fan-blowers is so slight 
that the lightest band-wheels and gears will answer, and when oper- 
a1 ed by hand only the slightest effort is necessary. The velocity needed 
is rapid, yet not exceedingly high, but should preferably equal one 
thousand revolutions or more per minute. The difficult part of this 
problem pertained to the devising of a practical method of feeding 
powder or fluid by a regular and gauged supply in small quantity into 
the blast in the encasement or spouts of such blowers, and to the pro- 
duction of suitable distributors to deliver the laden blast upward into 
the plants in a simple and efficient manner. With reference to the kind 
