488 TKOCEEPINGS OF THE THIRD ENTOMOLOGICAL MEETING 



giving the liumidit}'' record. The curve obtained has been re-drawn 

 on the humidity chart. One gramme of lint was used. 



Seed-weight is also a variable factor, but apparently changes with 

 a lag of a few hours, as can be seen from the second graph made in 

 the same way but with seed instead of lint. The third graph made 

 with two grammes of seed on one pan and one gramme of lint a one 

 gramme weight shows how humidity is constantly modifying the percent- 

 age lint. 



There is no correlation between seed- weight and percentage lint ; 

 on the other hand, the correlation between the lint produced per seed 

 and percentage lint is very high, as was to be expected. (We found 

 it to be 0-905 ±0013). The correlation between lint produced per boll 

 and percentage hnt is lower, being calculated from the same samples 

 at 0.367 ±0-064. 



These being the limitations of the percentage lint as a means to 

 gain evidence as to the damage done by the Pink Bollworm, we can 

 proceed to apply the method. And it may in advance be stated that 

 there is a loss of lint directly proportionate to the loss of seed substance, 

 involving practically no fluctuation of the percentage lint. 



The effect of GelecMa attack on seed cotton is to decrease the pro- 

 duction of seed and lint. For the seed we have already sho-mi a 

 decrease (1) in numbers matured, (2) in weight of the sound seed matured, 

 (3) actual loss of substance produced in attacked seeds. The effect 

 on the quantities of lint produced has been studied indirectly by exami- 

 nation of the percentages of lint. 



Obviously the suppressed seeds mature no lint, and are piue loss. 



It has not been possible to find any evidence that the reduction 

 in weight of sound seeds from attacked bolls is accompanied by any 

 change in the percentage of lint priJduced by such seeds. The loss of 

 lint in such sound seeds may be considered i^roportionate to their loss 

 of weight. 



The changes in the percentage of lint produced by damaged seeds 

 are more difficult to summarize, and depend to some extent on the age 

 of the seed at the time of the attack. 



(1) If the seed is attacked at a very early stage, it disappears 



entirely, thus causing the reduction observed in the average 

 number of seeds set remarked on above. 



(2) Attacked slightly later, part of the seed remains but sets no 



lint, thus causing the percentage to fall to per cent. 



(3) Attacked when nearing complete maturity, no damage is caused 



to the lint, but the seed loses substance, thus causing a 

 rise in the percentage of lint observed. 



