342 Bulletin 313 



It is clear that the most important problem before hay growers to-day 

 is to secure improved varieties that are known to be adapted to local 

 conditions and fitted to give the best quality and yields of hay under 

 these conditions. Not until good varieties adapted to various regions 

 have been secured can we expect any very general improvement and 

 increase in the hay crop of the State or of the country. 



METHODS USED IN THE CORNELL EXPERIMENTS 



During its many years of cultivation over extensive areas in various 

 parts of the world, timothy has been subjected to widely varying con- 

 ditions of climate and of soil, and if different environmental conditions 

 tend to cause or induce variations we should expect to find in timothy a 

 wide range of variations. In starting to improve any plant, the first 

 step in the investigation is to collect the known varieties and as many 

 different variations as can be found, and test them comparatively under 

 the environmental conditions where our experiments are to be conducted. 

 This is done in order to learn which ones are best suited to the particular 

 conditions of environment with which we are concerned. Timothy has 

 not been segregated into varieties, and the first stage of the breeding 

 investigation becomes thus the collection of as many variations as pos- 

 sible. The problem of the breeder here is to determine what different 

 characters have been originated in the plant in the course of its cultivation, 

 and then to find individual plants in which those characters that are most 

 desirable are combined. If such plants cannot be found already in exist- 

 ence, the breeder must resort to hybridization in order to obtain the 

 combination of characters desired. 



In starting the Cornell experiments with timothy, in order to determine 

 the range of variation and to procure material from which a selection of 

 types could be made, timothy seed was obtained from a large niimber of 

 places in the United States, Canada, and foreign countries. The various 

 sources from which seed was obtained and the number of samples from 

 each State or country were as follows: 



United States: 



New York 93 



Wisconsin 16 



Ohio 14 



Pennsylvania 14 



Michigan 6 



Maine 3 



Connecticut 2 



Delaware i 



