302 OR. L. H. GOUGH. 



The " Gemmaiza crop " figures have a greater claim to be accepted as accurate. 

 There is no possibility of errors in sampling ; the entire yield has been taken 

 under circumstances which exclude error as far as is possible ; the sound and the 

 damaged bolls were separated from each other carefully, and although a small 

 percentage of attacked bolls were included under the sound ones, the proportion 

 of attacked seeds wrongly included is less than 2 per cent. The damaged bolls 

 and the sound bolls together make up the entire yield of all the trees from which 

 they were taken. 



It will be seen that the only set of samples needing control and for whose sampling 

 no check can be found are the damaged seeds other than double ones. 



The following conclusions appear to be justifiable on comparing the percentages 

 of lint of all the samples together before removal of the larvae. 



The 30 sets of sound seeds approach very closely in their percentage lint to that 

 found for the entire crop, the difference being only - 2 per cent. ; the closeness of 

 results, in view of the small size of the sample, may in part be due to chance. 



The 30 sets of random samples are also not very far out in percentage lint, 

 beiug 0"7 per cent, in excess. 



The " damaged " seeds and the " double " seeds varied most from the normal, 

 as was to be expected, but it is almost surprising that they varied so little. The 

 " damaged " seeds were only H per cent, above the normal. Remembering that 

 the sound seeds varied to half this extent from the normal, and the entire bulk of 

 the 30 samples was in itself small, it is questionable how much significance should 

 be attributed to the figures. On the other hand, the " double " seeds are 3| per 

 cent, below the normal. In their case, part at least of the difference may be real, 

 especially as a certain amount of the lint on both components is probably destroyed 

 by the bollworm when attaching the seeds together. Hitherto, we have been 

 reckoning the weight of the larvae in with the seed-weight, as would happen at 

 a ginnery when calculating ginning output. Obviously the weight of the larva 

 compensates to a large extent for the lost seed-weight in the percentage lint calcula- 

 tions. However, on comparing the total weights produced by damaged seeds 

 and double seeds with that of the sound seeds, it is very obvious that a considerable 

 loss of substance has occurred, which strangely enough is distributed in such a 

 manner that lint and seed plus larva are in very nearly the same proportions as 

 lint and seed in normal seeds. Removing the larvae from these samples, the 

 percentage lint rises two to three per cent. 



The " Gemmaiza crop " samples are much bigger individually and together 

 than the sets just under consideration. Here, as was normal, the component 

 samples varied in percentage lint, the standard deviation being approximately 

 0'5 per cent. The entire samples on the other hand all worked out to 34"0 per 

 cent, lint (seed weighed including larvae). 



Even supposing the sets of 30 samples of " damaged " and " double " seeds to be 

 absolutely representative of the change produced in the percentage lint of such seed, 

 it may be remembered that changes of such magnitude would not occur in ordinary 

 samples, a 100 per cent, infestation of seed being very exceptional ; for 100 per cent, 

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



