234 HORTICULTURAL SOCIETY OF NEW YORK. 



but the losses have also been very great under this plan. The scientific way 

 is to do one thing at a time, and work from the simple to the fcomplex. In 

 this way I would suggest that ten or a dozen suitable species should be 

 selected for experimentation. 



One plot of individuals should be well cultivated and self-fertilized, 

 their seeds being again sown so as to get at the simple results of good 

 cultivation and selection alone. The plants in plot 2, under the same con- 

 ditions, should be carefully hybridized, reciprocally if possible, and the seeds 

 of these should be again sown and well grown. Selection might be made in 

 both cases, the object in view being to decide whether the simple selection 

 of self-fertilized seedlings does not play a larger and hybridization alone a 

 smaller part in the evolution of garden plants than is at present believed to be 

 the case. The experiments could then be continued with the same material, 

 so as to determine the importance of the part played by the cross breeding 

 of the selected varieties in both cases. As it is, we are in "going direct" 

 working with unknown factors. We must first of all find out how our parent 

 species behave under, i, culture; 2, selection; 3, cross breeding, and 4, hybrid- 

 ism, instead of hybridizing first and trusting to chance for our results. When 

 we see the wonderful results attained among live stock, cattle and poultry, 

 as well as among fruits, vegetables and flowers, by cross fertilization and 

 selection, we may realize that after all hybridism is not everything in the 

 evolution of the most useful animals and plants of both the farm and the 

 garden. 



