64 Cooperative Investigations an Plauts 



would have done under the Hanipden conditions with what the offspring plants 

 actually did in tlie giveii locality. 



The mean of the 706 capsules on tlie Hampden parent plants was 1275 with 

 a Standard deviation "f 1788, while the total Haiiijxlen crop had a capsule nican 

 of 12'51 and Standard deviation of 1"S98. Thus niy selected parentage had a 

 higher mean and a much reduced variability, however measiired. It will be 

 observed, however, that the weighted parental mean, i.e. parents weighted with 

 the unmber of offspring capsules, always differs from the mean of the pareut 

 plants. We cannot, however, deterraine how much of this was due to environ- 

 ment; the seed of the original 24 parent plants could not be distributed in] 

 normal population proportions, the offspring of each plant wliich survived differed/ 

 in number according to the amount of seed sown, the thinning, etc. At Chelsea 

 the storm chanced to destroy the offspring groups having parents near the mean, 

 and thus artificially exaggerat(3d the variability of the Chelsea parental selection. 

 At Oxford where \V. F. R. Weldon attempted to deal with individual plants 

 by sowing in pots (and so by subjeeting individuals to a like treatment a record 

 of tlie relative fertility of tlie parent seed was possible) it was found that even 

 the nattire of the ground immediatelj' under individual pots influenced individual 

 development. Thns at Chelsea and Bookham, only 12 and 16 of the parents 

 respectively contributed to the offspring, and this fully accounts for the varia- 

 bility of the actual weighted parents being in thesc cases considerably greater 

 tlian that of the original Ilampden parentage. The conditions at both Highgate 

 and Oxford were adverse to the plant; at Highgate one capsule to a plant was 

 about the avcrage, and at Oxford the average was only 32. The low variabilities 

 of the parentages at these places, are therefore probably due to the selection of 

 parent seed by the environmental conditions, but a uiore definite statement than 

 this it would be unwise to make. Comparing, however, the actual parentages with 

 the local offspring we see at once that the Bookham and Enäcld environments 

 were more favourable to the development of stigmatic bands, and Oxford, 

 Highgate and Chelsea less favourable than Hampden. The Kidderminster crop 

 while far heavier than the Enfield in the number of fruit per plant shows a 

 diminution in the mean number of Stigmata. Thus the environment largely 

 affects the number of Stigmata, but there is no evidence to show that this is 

 appreciably influenced by parentage in bulk. Tlius while the selected parentages 

 ränge only from 1204 to 1.S'21, a ränge of ri7, the resulting offspring means 

 ränge from 1076 to 1.'5!).5, or a ränge of 3'19. Nor is there any System in the 

 arrangement, e.g. 



Order of Parentage Order u( Oflspring 



Chelsea Oxford 



Bookham Chelsea 



Oxford Highgate 



Enfield I Bookham 



Highgate Kidderminster 



Kidderminster Enfield I 



