MEDICAGO FALCATA, A YELLOW-FLOWERED ALFALFA. 



21 



these lines can not be made. In order to include in the same class 

 the forms that most naturally resemble each other, it seems necessary 

 to make the divisions along the line of plant type Avith regard to 

 erectness, type of crown, and general habit of growth. A classifica- 

 tion even on this basis has in it many inconsistencies, not only with 

 regard to special botanical characters, but also in the general charac- 

 teristics upon which it is founded. Its weaknesses from a botanical 

 standpoint are fully appreciated, yet it seems to be less artificial than 

 the other classifications that suggest themselves. It possesses the ad- 

 ditional advantage of being based on characters of agronomic im- 



Fio. 8. — Individual plant of Medlcapo fnlcatti, S. P. I. No. 20725, as viewed from directly 

 above; a very proeiimbent, fine-Ktemined form, illustratins (iroup I. 



portance; in fact, it is jjei'haps moi-e agioiioiiiic than botanical, and is 

 certainly nun-e in the nature of u convenient grouping than an outline 

 of actual relationships. It is hoped that a further study of the mate- 

 rial will result in the develo[)m('nt of a more satisfactory classification. 

 The system which is adopted herein provides for four gi'oups. 

 Certain of these are pretty clearly defined, while some differ from 

 others only in degree;. Each group is illustrated by a figure of an 

 individual plant, and for the benefit of those to whom (he (lei)art- 

 inent's introductions of Afed/'oof/o falcata are availal)le S. 1*. I. num- 

 bers ai'c cited, which inebide forms (Inil ai'e tyi)ical of the gi'oups. 



