34 PROOFS OF HEREDITY [CHAP. 



natural, the contention is overwhelmingly estab- 

 lished that xerophytic characters are always 

 acquired by the soma or vegetative organs long 

 before there are any flowers with germ-cells, 

 and yet they prove to be hereditary, for a 

 shorter or longer time, until the environment 

 is changed again, and new adaptations arise 

 afresh ; at the same time they retain their 

 hereditary floral characters of affinities, which 

 themselves had been originally and similarly 

 acquired. 



Similar conclusions are arrived at by revers- 

 ing the conditions from water to air. I experi- 

 mented with a variety of the water crowfoot 

 (Ranunculus heterophyllus, fig. 5). Sowing 

 the seeds in garden soil, about one hundred 

 and fifty to two hundred seedlings appeared. 

 They were all alike, and first bore finely-dissected 

 leaves, as if they were under water, but the 

 entire anatomical structure was changed, having 

 a true epidermis and stomates, which do 

 not exist on submerged leaves. After a time 

 the completely formed floating type of leaf 

 appeared, as if the stems had reached the surface 

 of imaginary water, and, lastly, flowers were 

 borne in the air, as always is the case. This 

 experiment proves that the two kinds of leaf 

 were hereditary. I took some of the plants as 

 soon as the aerial type of the dissected foliage 

 was well developed, and transferred them to a 



