306 DR. L. H. GOUGH. 



random bolls of the same set plus the amounts found for reduction in average seed- 

 weight, plus the amount of the difference in actual weight of the damaged seeds from 

 the weight of the same number of sound seeds, plus the amount of weight of the missing 

 seeds after these three amounts have each had added to them 50 per cent, to 

 correspond to their lint- weight, at 33-3 per cent. lint. It will be seen that the 

 calculation is only once wrong to as much as 1 per cent, (which in this case is due to 

 differences in the percentage lint of the average sample and of the sound sample, 

 which were here greater than in the other pairs of samples, but to which no special 

 importance need be attached in view of what has been said earlier). 



From the close correspondence of the calculated totals of loss to the amount found 

 by weighing, it appears reasonable to believe that the percentage lint is not greatly 

 altered on account of the attack. Damaged seeds apparently produce lint on the 

 average in the same proportion to their remaining seed-weight (plus worm-weight) 

 as sound seeds from the same samples do, or else the differences between calculated 

 and actual weights would have to have been greater than found. (All seed-weight 

 is here given with the weight of the larvae included.) 



It is of importance to note that, as a general rule, the damage done to the attacked 

 seeds is only about half (or even less) of the total damage, and that it is often equalled, 

 and in some cases even exceeded, by the loss caused by the diminution of weight 

 of the sound seeds or the non-development of other seeds. 



It is further of interest that in these samples the average damage done has been 

 roughly proportionate to the percentage of bolls attacked, the loss being somewhere 

 about one-fifth of the total amount which could have been produced by the attacked 

 bolls. This would give about 20 per cent, loss if all the bolls in a sample were attacked. 

 However, it is necessary to remember that the total loss might be much greater when 

 100 per cent, of the bolls are attacked, as then the extra damage due to multiple 

 attack would make itself still more felt than it does in the figures under examination. 

 Some at least of the irregularities of the percentage loss figures are due to this cause; 

 with reference to which it may here be mentioned that the higher the percentage of 

 attacked bolls, the greater the probability of multiple infestation of single bolls. 



Finally, it may be stated that the damage done in this one experimental field was 

 between 11 and 16 per cent, on the first picking, and between 17 and 20 per cent, 

 on the second picking. It must also be observed that all this quantity would not 

 have been saved if the conditions existing previous to 1912 had obtained. In those 

 days Earias insulana did a certain amount of damage annually, an amount which 

 is stated to have been subject to considerable seasonal fluctuations. Balls mentions 

 18 per cent, damage as having occurred (Evading the Bollworni, Near East, 1915, 

 p. 332) in a bad bollworm year. As Gelechia had not become a major pest in his 

 time, and as Balls writes of bollworm and not of pink bollworm, he was j>resumably 

 referring to Earias. Now since the advent of Gelechia it has been very noticeable that 

 Earias has very nearly disappeared altogether as major pest. (In the samples under 

 discussion it was not foimd to have done any damage.) It is not supposed that Earias 

 has been suppressed by Gelechia, but the spread of the latter in Egypt has merely 

 happened to coincide with legislation which has made conditions unfavourable to 

 Earias. Further, it must be remembered that bolls attacked by Earias were 

 destroyed to a much greater extent than those attacked by Gelechia. A considerable 



