-10U- 
Several methods were used for extracting the "bugs from the samoles. 
(l) Direct dissection over a large sheet of oilcloth or paper, removing and 
counting the bugs as they emerged or were uncovered. (2) Direct dissection 
and sifting, first with a fine sieve to remove soil, then with a coarse sieve 
to remove the coarse litter, and finally counting and removing "bugs as they 
emerged or were uncovered from the remainder. (3) Use of Eerlese funnels 
under steam pipes to drive bugs by means of heat into collecting jars con- 
taining alcohol. (U) Submergence and dissection of sample in a tub of water 
to float out the bugs. For lightly infested samnles methods 1 and 2 were 
best, but for heavily infested sanroles methods 3 and U were preferable. 
Variation among Individual Samples 
The numbers of bugs in individual samples were extremely variable, 
even when taken from the same kind of grass and as nearly as nossible from 
the same location. For instar.ce> six l/5^-squc.re-foot end six 1-square-f - 
samples, all taken from big bluestem grass within 100 yards along a well- 
drained southward sloping roadside, varied as follows: 
Size of sample - 
Bus:s tier 
sample 
M i n imum : Max imum 
Average 
Number. Nur.be r 
Number 
1/5 square foot 
1 square foot-- 
lU : 236 
232 : 3 , 169 
107 
916 
Further evidence of the extreme variability in numbers of bugs in 
individual samples is given in the following comparison between l/5-square- 
foot and 1-square-f oot samules. Twenty-five nairs were taken from the most 
favorably located clumps of big bluestem in a fairly uniform area about 2 
miles square, each pair consisting of adjacent l/5-square-foot and a 1-square- 
foot samples from the seme clump of grass, with the following results: 
Size of sample ' 
BU£8 
per sample 
Minimum 
: Maximum : 
Average 
l/5 square foot — 
1 square foot 
Number 
1 
■ U 9 
: Numb or 
: 2,563 
: 3,169 
"umber 
: 3^0 
• "676 
Notwithstanding efforts to take the above samples from as uniform an 
environment as possible the variation between those of either size was very 
great. The data were analyzed by the usual statistical methods but the re- 
sults are not given here because the methods of analysis commonly used were 
apparently not applicable to this experiment. The distribution of chinch 
bugs in hibernation is so uneven that a great many more and also much larger 
samples, perhaps entirely on an area basis, irrespective of host plants, 
would probably be necessary before the data would abroach anything like a 
