The Cultural Evolution of Cyclamen latifolium. 147 



sponding rate in the organisms which then existed."* That changes 

 may be effected with considerable rapidity cannot, I think, be 

 denied. 



2. It is further, I think, abundantly proved in the present case 

 that, though sudden variations do occur, they are,, as far as we know, 

 slight as long as self-fertilisation is adhered to. The striking 

 results obtained by cultivators have been due to the patient accumu- 

 lation by selection of gradual but continuous variation in any 

 desired direction. 



3. The size which any variable organ can reach does not appear to 

 be governed by any principle of correlation. Large flowers are not 

 necessarily accompanied by large leaves. Under natural conditions 

 size is controlled by mechanical limitations and by the principle of 

 economy. Nature cannot afford to indulge in anything unnecessary 

 for the purpose in view.f 



4. The general tendency of a plant varying freely under artificial 

 conditions seems to be atavistic, i.e., to shed adaptive modifications 

 which have ceased to be useful, and either to revert to a more gene- 

 ralised type or to reproduce " characters which are already present 

 in other members of the same group."! This conclusion must, 

 however, be accepted with caution, for we must remember that in a 

 case like the present we are only acquainted with variations which 

 have been preserved with a particular end in view. 



5. The case of " cresting" shows that the plant still possesses the 

 power to strike out a new line and to develop characters which 

 would even be regarded as having specific value, as in the total 

 change which has been effected in the form of the leaf in Primula 

 sinensis. If such a race developed any degree of sterility with other 

 races it would have satisfied Huxley's criterion for the artificial 

 production of a new species. 



* ' Origin,' 6th ed., p. 286. 



f See Darwin, ' Origin,' 6th ed., p. 117. 



X See Darwin, ' Origin,' 6th ed., p. 127. 



VOTy. LXI. 



M 



