22 JOUEXAL OF THE MiTCIIELL SoCIETY [July 



lowlands, plants that had grown at elevated locations for four 

 to six years retained their alpine characters for four to six 

 3^ears, but eventually returned to the habit and character of 

 lowland types. 



The well known German botanist, Klebs, likewise ex- 

 perimented extensively in this line with the net result that 

 many acquired characters are inherited while others are not. 

 Perhaps experiment should now be directed toward ascertain- 

 ing just what acquired characters are inherited. 



Lest you should imagine that I have ignored the evidence 

 from the zoological side let me briefly refer to Kammerer's 

 results. It should be added, however, that zoologists are more 

 nearly a unit tlian botanists in the view that acquired char- 

 acters are not inherited. 



Kammerer at Vienna by reducing the amount of water 

 succeeded in permanently modifying aquatic species of sala- 

 manders so that they came more nearly to resemble land 

 forms and by changing the color of the soil on the bottom of 

 the aquaria transformed the color of the animals so as to 

 correspond very nearly with the color of the substratum. 

 These effects according to Kammerer are hereditary. There 

 are numerous other experiments of this kind made on ani- 

 mals with corresponding results, still it must be said that the 

 bulk of the evidence furnished by experiments on animals 

 seems to support the negative side of this proposition. Un- 

 fortunately, however, this is not a question that can be settled 

 by majorities in the good old democratic way. 



Castle, of the Harvard Zoological Department, as well as 

 several others, has made some beautiful experiments in the 

 hope of solving the prolilem. Their method consists in graft- 

 ing the ovary of one animal into another from which her o^v^l 

 had ])cen rcinovcd. The body of the animal operated upon 

 may be looked up(m as a new environment for the engrafted 

 ovary, and by reason of the intimate relations existing ought 

 to exercise a marked influence in heredity, if such a thing 

 be possible. 



The details of one of these experiments is as follows: 



