THE COXriirHLA. 59 
but later it was shown that data of this kind obtained at th.it season 
of the year were of little significance. Thus in one locality as high 
as 70 per cent of all bolls in July were ruined by plant-bugs while 
later increases in the number of bolls and decreasing destructive] 
of the insect lowered the percentage of damaged bolls to about 31. 
Method of estimation of losses. — Final estimates of the losses due to 
plant-bugs at Tlahualilo were made during the first week in Decem- 
ber, 1905. Growth of the bolls had practically ceased, although no 
frost had occurred. Many green bolls had attained a mature size 
and in all probability would open under the influence of the first 
frosts. Estimates for each section of a field were based on exact 
counts and the classification of all bolls on 25 plants of a row, selecting 
5 consecutive plants at each end and 3 groups of 5 consecutive plants 
each, at points between the end groups, to equitably represent the 
entire row. The bolls produced by each plant were classified as 
unopen, empty burrs, perfect bolls, slightly stained, badly stained, 
and destroyed. Included in this estimation as unopened were only 
such bolls as had reached mature size and were likely to open and 
produce good lint when not injured by insects. "Empty burrs " indi- 
cate that from these the lint matured and was picked or had fallen out. 
In either case as far as this investigation is concerned these should 
be associated with the perfect bolls. Classed as perfect bolls were 
only those which showed no trace of noticeable stain due to plant- 
bug attack. Included as slightly-stained are those which plainly 
show stain, although probably worth picking in the sense that the 
increased bulk gained thereby would probably offset the possible 
decrease in value per pound due to the stain. "'Badly-stained" 
bolls contained no lint that could be profitably picked, although not 
more than one lock in any boll or about 10 per cent of all locks in 
this class was actually destroyed. '"Destroyed bolls" included no 
lint that a picker would often intentionally pick, and were char- 
acterized by open or partly-open carpels showing a discolored, matted 
mass of partially developed lint in at least two locks of each boll, 
with any locks not thus affected badly stained and damaged suffi- 
ciently to pi-event "blowing." Classification as regards the Last four 
grades was in many cases necessarily a matter oi personal opinion, 
but in averaging it is believed this feature is largely eliminated, as on 
the whole the differences are quite distinct. Bolls damaged by boll- 
worms are omitted from these considerations. 
Damagi to cotton in. Ceceda A. No. / .;. Five examinations were 
made according to the methods described above. One was about 
50 feet from the east end, 1 about the same distance from the west 
nn\, and 3 at points dividing the block into 4 approximately equal 
