1902.] I. H. BmkiW— Flower of Ranunculus arvensis. 



Table III. — Variation in Stamejis. 



99 



No. of Stamens. 







1 



2 



3 



4 



5 



6 



7 



8 



9 



10 



11 



12 



13 



14 



15 



1617 



18 



Heidelberg 





1 



12 



21 65 



136 



238 



308 



282 



170 



98 



32 



8 



4 



3 



3 



1 



] 





Bonn 





] 



4 



9 21 



81 



154 



247 



294 



211 



123 



36 



8 



5 



6 







2 



1 



... 



Kew, Old ... 



... 



1 



B 



62 229 



433 



428 



304 



270 



134 



140 



97 



73 



69 



27 



17 



9 



... 



., 



Kew, Now 



... 



3 



6 



30143 



292 



265 



203 



190 



140 



95 



74 



60 



46 



25 



1 



4 



1 



1 



Table IV. — Variation in Carpels 



No. of Carpels. 







1 



2 



3 



4 



5 



6 



7 



8 



9 



10 



n 12 



,3 



14 



15 16 



Heidelberg 

 Bonn ... 



3 

 5 



3 

 10 

 3 

 2 



13 



4 



24 



16 



95 



35 



175 



68 



266 

 155 

 362 

 198 



511 

 426 

 634 

 353 



314 

 379 



584 

 391 



136 

 153 

 403 

 364 



81 



28 

 107 

 158 



8 



6 



6 



37 



r 



2 







1 











1 



1 



Kew, Old 















Kew, New 



1 



' 



























There is an obvious difference between the two German races and 

 the Kew race and some difference between the Kew plants from 1896 

 seed and those from 1897 seed although they belonged to the same stock. 

 The variation curves which may be plotted from these figures are irre- 

 gular, and those for no one set of organs exactly correspond with those 

 for neighbouring sets : the curves of the sepals are half-Galton curves : 

 and the curves of the petals in the Kew race are also half- Gal ton curves, 

 but not quite as those for the sepals ; while the curves of the petals in the 

 German races are intermediate between half-Galton and symmetric 

 Quetelet binomial curves : the curves for the stamens are equally asym- 

 metric, but in a different way ; while the curves for the carpels are the 

 most nearly bi-symmetric of all but are not quite so. It is evident from 

 a comparison of them that the flower does not vary as an unit as for in- 

 stance a Tulip flower may, every ring of organs changing from 3-merism 

 to 4-merism ; but each mood varies in its own manner. We shall learn 

 more of this independence of the moods in variation by studying their 

 association. I cannot give tables of the combinations observed in the 

 different races for all the four sets of organs taken two together, without 

 occupying a great amount of space ; I therefore give tables for the 

 " Kew Old " plants alone. They will serve as an illustration for all, 

 as the tables which could be given for the German races and " Kew 

 New " are not unlike them. 



