TESTS. OF MACHINERY FOR KILLING COTTON WORMS. 45 
sary for thispurpose. The field laborer of the Southscrews up the joint 
too tight, too loose, or in such form as to spoil the screw-threads. Again, 
the joints become rusted together and a vise must be engaged. The 
stiff system also requires that very heavy pipe be used, as the leverage 
on long pipe arms enables them to suffer great strain, to become broken 
off easily at the end where the thread for the joint is eut, whereas with 
flexile joints no leverage power but only tensile strain can be brought to 
bear. In the latter case very light tubing can be employed with economy 
in material, cost, labor, and salvage of cotton. Moreover, only by such 
light flexile apparatus can any considerable number of rows be treated 
at once from beneath. These facts have been substantiated by tests of 
stiff and of flexile apparatus this season more fully than they were by 
the Atlanta tests, in which, one ight machine undersprayed eighteen 
rows of cotton, a strip twenty yards wide, at a single drive. The tests 
this year have been not only of stiff connections, but also of the con- 
structions whereby adjustability of the descending pipes is effected au- 
tomatically and by hand. These have already been noticed above or 
in the previous reports in so far as they pertain to the stem or body of 
the pipe or its distal appendages ; hence, next in order may be considered 
more specifically and in natural sequence the construction and arrange- 
ments of the tops of these pipes as planned and tested by me: 
I. The stiff hanging tubes have been tried, as already set forth, in 
firm union with a stiff back-pipe or cross-pipe such as appears in many 
of the patented sprayers, as Johnson’s, Daughtrey’s, etc., while suffi- 
cient objections to this arrangement for underspraying have already 
been presented. Itis the first construction which naturally suggests 
itself to any plumber or other mechanic, but presents no special adap- 
tation for the purpose, as has been shown this season and previously. 
Il. The extremest opposite construction to the foregoing is attained 
by having radiating flexile tubes from the main to the descending pipes, 
instead of a straight and stiff cross-pipe. By this arrangement the 
hanging pipes are swung apart or nearer together independently, and 
set on a cross-bar or on diverging bars, at spaces to suit rows having. 
different courses or widths. 
By way of variation the tubes may radiate only for a part of the 
distance, and for a space run close beside each other along a supporting 
bar before reaching the descending parts. The parts upon the support 
are preferably of metal, and slide readily in peculiarly locked hooks, as 
simple, easily separable attachments, specially devised for this purpose. 
Where the descending parts have flexibility to some extent they may 
drag in the cotton in turning,as stated above. It is shown that they thus 
dono noteworthy harm to the plants; also that they themselves do not 
sufferinjury. -This flexile construction is simple, and generally preferable 
in combination with the flexile connectives between their tops. But 
should any prefer that the hanging parts be elevated above the plants in 
turning, this iseasily done. For such purpose, and to shorten the lever- 
