6 BULLETIN" 121, U. S. DEPARTMENT OF AGRICULTURE. 
DIFFERENCES IN LENGTH OF STAPLE. 
These tests were designed to determine approximately the relative 
amounts of waste in the several varieties and the breaking strength 
of yarns made from them. The Delta was included so that compari- 
son of results could be made not only with the ordinary commercial 
calculations for lj-inch cotton similarly handled, but also with actual 
tests of such cotton made on the same machines. Each of the live 
bales tested in this experiment was run on the same machines and 
with absolutely the same settings, so that the data were strictly com- 
parable in every way throughout. Notwithstanding the fact that 
the staple of the different samples varied from 1 T 3 6 to 1 T % inches, 
they were all put through the machines with the settings used for 
lj-inch cotton, because each of the varieties tested is believed under 
favorable conditions to be in competition with lj-inch Delta staples. 
The Columbia and Durango were apparently at a disadvantage in 
this test, because with their shorter staple the} 7 would naturally give 
ar- increased percentage of waste. Table I shows, however, that the 
total picker and card waste was very nearly in proportion to grade, 
the Lewis apparently being slightty better than the others. 
THE INVISIBLE LOSS. 
The invisible loss, which is shown in Table II, is caused chiefly by 
loss of moisture and also to some extent by dust and short fibers 
which are lost in the air during the operation of stripping at the 
card. There is no appreciable invisible loss after the cotton has 
passed through the card, as all dust, short fly waste, and excessive 
moisture have been removed. The percentage of humidity main- 
tained affects the amount of invisible loss, as the greater the humidity 
in the mill the more moisture will be retained in the cotton within 
certain definite limits. There being no humidifying apparatus in 
the picker, card, and comber rooms at the Xew Bedford Textile 
School, the humidity during these tests could not be regulated. It 
will be seen in Table II that the invisible loss is greater in the low 
grades, probably because they contain more dust and light trash. 
Table II. — Percentage of invisible loss at each machine in the picking and card- 
ing processes. 
Breaker picker 
Intermedial" pi kw. 
Finisher i>i< ker 
Card 
Total. 
Variety. 
Durango, 
grade 
G. M. 
Columbia, 
grade 
8. G. M. 
Lewis, 
grade 
<;. m. 
Webber 
No. l. 
grade m. 
Webber 
No. :'. 
grade M. 
1.875 
.58 
.071 
.503 
0. 37.", 
.77 
.144 
1.31 
1.125 
.38 
.49 
.48 
1.75 
.7s 
.31 
1.49 
2.00 
..".7 
.72 
3.029 
2.599 
2.4 7.') 4.33 
3.a 
Delta 
blend >d. 
(') 
1 No Invisible loss is shown for the Delta cotton up to the card, as this sample was started at that machine. 
This blend contained one strict low middling, two middling, and two good middling bales. 
