I. I iiheritance in Shirley Popp]i 83 



(ii) Local ciivironim'iit immensely influeiiced the vuriabilily and incaii ot' tho 

 looal crops. Tliis lias. howcvcr, lit-tlc or im iiiflucnce on the dclciiiiinaticiu ot' 

 hc'ivdit.y. What docs iiiriueiice licredity i.s differentiation in the loeal eiivironinunt 

 of a crop. In such a case when f'cw parents aro soloctud to start with a dit'i'cR'ntial 

 treatnicnt ot' thosc individiial parunls iiiay nuich modif'y results, or again a 

 diff'erential troatuient of the sub-crops inay be very detrirneiital to oonsi.stent 

 results. Soniething of this kind seems to have afteeted the Kidderminster crop, 

 for paroutal heredity is woakened and fraternal exaggerated — just what we should 

 expect from thi' lattcr fdiin of local environuient difl'erentiation. 



(iii) Most plant organs being multiple in appearance, we have to apply 

 special niethods to deduce the intensity of heredity frora multiple observations on 

 the individnal. It will probably be better in future experiments to confine 

 attention to the fir.st, or principal flower, instead of using the indirect method of 

 homotyposis, but this will involve the Observation (previous to harvesting) of 

 individual plants in large series — 500 to 1000 — and much increase the labour of 

 superintendence and Observation. 



(iv) Notwithstanding the difficulties referred to above we find that for a 

 variety of plant characters in the Shirley Poppy the values of hereditary influenae 

 found are on the whole iu fair agreement with the like values for man. 



Undoubtedly tlie most .self-accordaut results are from Weldon's Oxford crop, 

 where growing in pots, although it tended to produce starvelings, gave owing to 

 the careful mixture of soil and administration of water, etc., a probably greater 

 equality of individual environment. 



Here tbc parental heredity was '4 and fraternal heredity •25, results in good 

 accord with those for man. The Highgate results, on incomparably more meagre 

 data, of about "5 and 25 respectively are also in good agreement. Enfield and 

 Kidderminster are by no means so satisfactory. Yet when we pass from stigmatic 

 band to colour inheritauce at Entield, we reach a good value, of '41, for parental 

 inheritance, and for all cases our parental meaus ('33 to ■34) are not below Francis 

 Galton's first determination of heredity in man. 



Hence these, the tirst observations on large series of the laws of inheritance in 

 the coutiuuously varyiug characters of plants, show numerical results generally in 

 accordance with those already found for animals and insects. 



The great influenae of environment, and of local differences of envii-onment, 

 the probable stringent selection of seed and seedlings, the fewness of the original 

 mother-plants, are all factors tending to modify and obscure the numerical results ; 

 they make the plant-problem much harder tlian the animal or insect problem. 

 But the present investigation shows that there is nothing, which would lead us to 

 suppose that the methods and results already found sufticient to describe hereditary 

 influence in man and animals will not suttice to describe the like results in 

 the continuously varyiug characters of plant life. 



11—2 



