J'KOCEEDlIs-G.S OF THE THIL'D ENTOMOLOGICAL MEETING 491 



Even supposing the sets of 30 samples of " damaged " and " double " 

 seeds to be absolutely representative of the change produced in the 

 percentage Hnt of such seed, it may be remembered that changes of 

 such magnitude would not occur in ordinary samples, as 100 per cent, 

 infestation of seed is very exceptional. 100 per cent, infestation of bolls 

 does not necessarily mean more than 6-6 per cent, infestation of the 

 seed, and rarel}^ more than about 20 per cent, (Crop conditions 1917).. 

 The alteration to the ultimate percentage lint would also not be in the 

 ratio percentage infestation x abnormal percentage lint : percentage 

 sound seeds X normal percentage lint, but would be altered in the direc- 

 tion of the normal percentage lint by the less weight produced by the 

 damaged seeds. 



The differences in direction of variation of percentage lint of 

 double and ordinary damaged seeds would also tend to eliminate each 

 other. All these factors together help to keep the percentage lint close 

 to what would have been the normal, if the Pink Bollworm had been 

 absent. 



In passing it may be remarked that the ginning outturns published 

 for the last few years have not showTi any fluctuations which can be 

 traced to damage done by Gelechia. (Plate 91). 



Tables XIV and XV have been compiled (from figures supplied by 

 the State Domains Administration, to whom our thanks are due) to 

 show how far the percentage lint in Mitafhfi and Afhfi has varied during 

 the period 1891-1917 and in Affifi, Assili and Sakellaridis in the period 

 1911-1917. 



The figures for Mitafiifi and Affifi can be considered comparable 

 to a great extent. The mean percentage lint for Mitaffi£ is 34-0, for 

 Afiifi 34-1 ; the two series are consequently extremely suitable for 

 comparison and can be considered as forming an unbroken series. 



The first point that calls for remark is that for every year during 

 the whole period Qorashia ginnery has never given a lower percentage 

 lint than Sakha ginnery, and in all but two years it has given a higher 

 percentage. 



The second point is that the simultaneous difference between the 

 percentages obtained by the two ginneries is often quite as large or 

 larger than the yearly fluctuations from the mean percentage. 



The third point is that the standard deviation for Mitaffifi (period 

 1891-1912) is twice as great as that for Affifi (period 1911-1917). As 

 Mitaffifi was grown previous to the introduction of Gelechia, and Affifi 

 is contemporaneous with that pest, it is evident that the percentage 

 lint has not been greatly influenced, if at all, by the Pinlv Bollworm. 

 The reduction of the magnitude of the standard deviation however,. 



