The Production of New and Improved Varieties of Timothy 359 



average yield per plant of the progeny of the same plants from open- 

 fertilized seed, each case being the average yield per plant of about 

 twenty-four plants for the second and third crop seasons of their growth. 

 In the fifth column is given the average yield per plant of the progeny 

 from self-fertilized seed, in each case being the average yield per plant 

 of about twenty-four plants for the second and third crop seasons of their 

 growth. 



It will be seen from an examination of the table that in the majority 

 of cases the original plant gave a larger average yield than did the clons 

 of the same plant ; the clons a larger average yield than the open-fertilized 

 seed plants ; and the open-fertilized seed plants a larger average yield than 

 the self -fertilized seed plants. The average yield per plant for all types 

 tested was: for the original plants, 9.307 ounces; for the clons, 8.769 ounces; 

 for the open-fertilized seed, 6.963 ounces ; and for the self -fertilized seed, 5 . 243 

 ounces. The open-fertilized plants thus gave an average yield for all 

 cases of 1.72 ounces per plant more than the self -fertilized plants. In 

 two instances the yield of the selfed-seed progeny exceeded that of the 

 open-fertilized progeny, and doubtless such cases m.ay be expected to 

 occur occasionally. While the general loss in vigor from inbreeding is 

 considerable, ordinarily not more than one inbreeding is necessary in 

 order to obtain a fairly uniform type, and when open fertilization is allowed 

 in the type later it is thought that the original vigor will be regained. 

 This point, however, has not yet been thoroughly tested. 



In making the first selections it was feared that the reduction in vigor 

 from inbreeding might prove a serious obstacle in the prosecution of the 

 experiments, and the scheme was tried of attempting to choose unrelated 

 seedlings exhibiting the same characters and general type, which could 

 be crossed together in producing a fixed type. In such tests the two 

 plants of the same type were planted close together in short clonal rows 

 so that the entire plat could be covered by a tent during the blooming 

 period, thus allowing the plants to interbreed freely. All such trials were 

 failures, as in every case the two types selected proved to be of different 

 germinal constitution and the progeny showed no general similarity of type. 



TESTING THE TRANSMISSION OF TYPE IN SELECTIONS WHEN GROWN 



FROM SEED 



(See Plates VII and VIII) 



In 1908 and 1909 test rows of plants from inbred seed of a considerable 

 number of different types were planted, and in many cases by the side 

 of these were grown, for comparison, similar rows from open-fertilized 

 seed and from clons of the same original plant. Here, as in all other 



