232 REPORT 4, UNITED STATES ENTOMOLOGICAL COMMISSION. 
an upward direction. The method of feeding tbe powder to the blast 
looks as though it cannot prove satisfactory. 
" Some suggestions made by the inventor as to the use of his machine 
for applying vapor and fine spray of liquid poison will be found men- 
tioned further on." * 
Rotary blowers of liquid. — In endeavoring to produce rotarj 
fans for blowing poison fluids the most difficult part of the problem 
pertains to feeding them to the blast. Since a blower will throw out 
any amount of fluid that is allowed to flow in, and this as fast as it 
enters, the quantity admitted must be limited to only so much as will 
make a spray of the quality and size desired. Also, tbe less fluid the 
finer and farther will it be blown, and if the supply is too great it will 
be dipped out or splashed in great masses or very large drops, which 
are not thrown far but wasted. To regulate or control this three meth- 
ods prove applicable. The first is by means of a rotary brush, as al- 
ready set forth in speaking of brush sprays. The second consists in 
reducing the orifice of the feed-pipe from the reservoir to such size 
that it will conduct just so much as is desired and no inore; that is, by 
means of a shut-off device or a small drip, by which the water enters 
the drum above so as to fall upon the revolving fans, or at some other 
point to accumulate in the bottom of the drum and be dipped up by 
them. The third principle is that employed in automatic inkstands, 
but applied on a larger scale. The reservoir is air-tight, closed by a 
can screw or plug. Its outlet is near its base, and extends through a 
tube leading into the bottom of the drum, where the fluid can flow in 
only just deep enough to shut off the entering of air into that end of 
the tube, while air will be admitted and the fluid pass out only so fast 
as the latter is lifted out by the fans. 
To prevent each fan from scooping up too much, also to divide the 
water mechanically and ventilate it thoroughly, the dipping ends should 
be provided with long comb like teeth. The best results are secured 
by bending each alternate tooth forward or backward at its base, thus, 
to make a double row, and then bend forward nearly at a right angle 
the distal half of each of all the teeth. This arrangement of the teeth 
is shown in Plate XXVII, Fig. G, i 7 and is explained below. 
The fans, by their rotation, give the fluid a tangential momentum, 
but the air current is a very important element, as I have shown by 
using toothed reels without fans on them. They are only equivalent to 
rotary brushes, which are here surpassed by introducing the superadded 
blowing power of the fans. 
The outlet from the drum should be large. In case it be low down 
upon the side, many drops will fall from or near it and the water may 
* To avoid much repetition of reference to Bulletin U. S. Entomological Com- 
mission, tin; first edition of this work, Prof. Barnard's credit-references to said 
Bulletin are omitted, so that all quoted matter, unless otherwise credited, is taken 
therefrom.— C. V. R, 
