(a) those containing to 4 worms, that ripen and produce pickable 

 cotton ; (6) those containing five to some higher number of worms, 

 that produce mabruma (i.e. unpickable cotton) ; and (c) those that 

 contain more than this higher number and which fall or dry up, and 

 are consequently not included as mabruma. 



Note. The reason for supposing four worms as the limit for 

 pickable cotton will become apparent later. 



(2) The number of worms may be limited by the attacked boll 

 becoming unsuitable for food for young larvae, which try to feed on 

 the boll and die from the effect. It does not appear unreasonable to 

 suppose that the boll gets badly contaminated with toxins arising 

 from the vital processes of the worms or produced by the plant in 

 response to the very serious injury done to it. The problem is one 

 for the physiologist not for the entomologist to investigate. The 

 result would require the production of the same number of worms 

 as/ giippiosed under the previous assumption, but (b) and (c) would 

 be mabrumfl., whereby the number of worms apparent in the (c) bolls 



: '.', ; ;' would; nci. exceed the top limit for (b). 



(3) The number of worms per boll may be limited by selection on 

 the part of the young worms, which do not enter bolls containing more 

 than a limited number. This would again have the effect of combining 



(b) and (c) as under the preceding paragraph (2), with the difference 

 that the larvae which had rejected heavily infested bolls would be 

 available to attack other less damaged ones. This would not alter 

 the method of calculation, except for the total production of worms 

 required to produce any given percentage. The total number present 

 whose activities could be demonstrated (supposing those that die 

 early evade observation) would be the same, but the number actually 

 generated from eggs would be less. 



The number of worms present in pickable ripe bolls remains the 

 same under all three assumptions, as does the number likely to be 

 observed in the mabruma bolls. Assumption one requires a reduction 

 of the apparent boll population, but not necessarily of the worm 

 population. Assumption two postulates reduction of the apparent 

 worm population, but not necessarily of the boll population. Assump- 

 tion three demands a higher percentage boll infection for any given 

 number of eggs laid, once certain limits have been surpassed. 



Should assumption one be correct, it would require (once the 

 number of worms causing boll-shedding had been reached) that the 

 true percentage attack should no longer be calculated from samples 

 actually gathered from the trees, as such samples would not contain 

 the fallen bolls. In all three cases, samples of "pickable" cotton, not 

 containing mabruma, cannot be used directly to find the true percent- 



