NOMENCLATUKE AND CLASSIFICATION. 25 



seeds are either identical or distinguishable with great difficulty. 

 Nevertheless, the results secured with other varieties leave no question 

 that all these can be separated and bred true to type. 



In regard to the brief descriptions given, a few words of explanation 

 are necessary. Many of the importations proved to be impure lots of 

 seeds. In some cases, especially where the seeds were differently 

 colored, these were separated before planting, and such are definitely 

 indicated. In other cases the mixture was not detected until the 

 plants were grown, or, in a few cases, until the seed was harvested. 

 Where the difference was detected in the field and the plants sep- 

 arated, they are referred to as "field selections." On the other hand, 

 if the selection was merely a separation of seed from the garnered 

 crop, these are spoken of as "seed selections." Both the "seed" 

 selections and the "field" selections are for the most part "mass" 

 selections, and many of them prove still to be impure, containing 

 both tawny and gray-haired, or red-flowered and white-flowered 

 varieties, which, however, mature together. Most of these have not 

 been separated, though in all valuable varieties they should be. 

 Where one or the other of such dift'erences is not recorded, the variety 

 is a pure strain. Where the selections were made the first year that 

 the plants were grown from imported seeds, they may be either acci- 

 dental admixtures or the result of hybridization at the place where 

 the original seed was grown. If, on the other hand, they were 

 selected two or more yearl after they were introduced, they are almost 

 certainly the result of hybridization at the Arlington Experimental 

 Farm. 



Besides these, many individual or centgener selections have been 

 made; these, however, are not considered in the accompanying de- 

 scriptions. Except these last, all selections are indicated by the origi- 

 nal S. P. I. serial number with a letter added, thus 16790 D. 



It will be apparent from the descriptions that many varieties are 

 very similar to one another. Only a comparatively few of them 

 have been named. Very careful field comparisons were made, how- 

 ever, in all cases, so that each description represents a different thing. 



In the cases of a number of early S. P. I. introductions, new num- 

 bers were assigned to different lots of seed grown from the original. 

 Thus, the original introduction of Ebony was S. P. I. No. 6386 and 

 different lots of its progeny were Nos. 8492, 9414, and 17254. This 

 is indicated in each case. Many of these earlier S. P. I. numbers were 

 also distributed under a series of Agrostology numbers, full keys to 

 their respective identities being given by Ball in Bulletin 98 of the 

 Bureau of Plant Industry, so that their identity with the numbers 

 and descriptions here given can be easily determined. 



197 



