354 Bulletin 313 



for further trials. Illustrations of what are believed to be good types 

 are shown in Plate II, Figs. 2,3, and 4, and Plate III, Fig. 6. 



Finally, the nutritive qualities and digestibility of the different types 

 should be taken into consideration, but in the first stages of variety for- 

 mation in a crop such as timothy these factors may well be neglected. 



TESTING SELECTED PLANTS OF DIFFERENT TYPES 



After a careful study of the variations and the different types 

 presented, a large number of plants were selected as representing interesting 

 and promising types. The important question then was, whether 

 the various individuals showing valuable characters would reproduce 

 those characters, or whether they were hybrids and unstable in nature. 

 The selection and testing of such variations has formed the basis of the 

 experimental work since 1907. 



A study of the progenies grown from open-fertilized seed of plants 

 selected in 1905 for special characters was found to show great variation 

 and little indication of a transmission of the characters for which the 

 plants were selected. Timothy is normally a cross-fertilized plant, and 

 when plants are grown in a mixed field, where they are freely crossed 

 with all sorts of pollen, this lack of transmission w^ould be expected. In 

 1907 a few heads protected from cross-pollination by covering with paper 

 bags gave a few good seeds, showing that at least some seed would set 

 by self-fertilization. Since that time, each season a considerable number 

 of the selected plants have been protected from cross-pollination by 

 covering the entire plant with small cloth tents, or by covering certain 

 heads with paper bags. In this way self-fertilized, or inbred, seed for 

 testing has been obtained from a large nvimber of select types. 



In general, the most satisfactory method is to cover a group of about a 

 dozen heads with a twelve-pound manila paper bag, just before they 

 begin blooming. In order to insure the pollination of the heads inclosed 

 in such bags, they should be shaken thoroughly early each morning during 

 the blooming period to aid in the distribution of the pollen over the pistils. 



The test of any particular chosen plant is conducted in the following 

 manner: 



I. The selected plant is propagated vegetatively by digging up and 

 separating the bulbs that are formed in the stooling of the plant (Fig. 91). 

 These are taken in early September and a row of sixteen to twenty-four 

 plants grown. These plants, it will be understood, are only transplanted 

 parts of the same individual. From such propagation the character of 

 the individual can be judged better and a more nearly correct idea can 

 be obtained of the yielding capacity of the plant as well as of other char- 

 acters. 



