"Lost" Species of Kentucky Liliaceae 53 



The problem concerning the existence of this species in Kentucky 

 today is one of the most interesting at the present time. Field work 

 is planned in Todd Co. in the near future in order to help answer the 

 question of its occurrence in the state. 



Distributional data given for Melanthium virginicum L. in Small 

 (1903, 1933), Robinson and Fernald (1908), Britton and Brown 

 (1913), Fernald (1950) and Gleason (1952) leave no doubt that the 

 species occurs in Kentucky. It is of significance, however, that all of 

 these reports are apparently based upon one collection, that of C. W. 

 Short. No others have been located in herbaria. Whether this species 

 still occurs in the state is problematic since the one known collection 

 was made 120 years ago. Deam (1940) gives Floyd Co., Indiana, as 

 one locality of this species, and since this county is across the Ohio 

 River from Louisville, this information should be of some help in 

 attempting to relocate the species. 



The following specimens are cited as the Kentucky records: 



( ? Co.). 1842. C. W. Short, M.D., s.n. (Two sheets, NY) 



Another species of Liliaceae which has been collected in the state 

 only once and which has not been reported since is Xerophyllum aspho- 

 deloides ( L. ) Nutt. This species has a very scattered distribution in 

 the eastern United States, and it appears to be found mainly in dry 

 habitats of the pine barrens of the Coastal Plain in "New Jersey, Dela- 

 ware and North Carolina" and in similar situations on mountain sides 

 and mountain tops from "Virginia to Georgia and Tennessee" (Fer- 

 nald, 1950). Its occurrence in south-central Kentucky is somewhat of 

 a puzzle in plant distribution, and in spite of attempts to do so, it 

 has not been possible to relocate it. Based on both floral character- 

 istics and length of the pedicels, the specimen cited below can only be 

 X. asphodeloides. Its only known occurrence in the state is repre- 

 sented by the following specimen: 



(Warren Co.). Bowling Green. 1903. Sadie F. Price, s.n. (MO) 



This species might most reasonably be expected to occur in the 

 Cumberland Plateau Province of eastern Kentucky since Robinson 

 and Fernald (1908) and Britton and Brown (1913) describe its dis- 

 tribution as: ". . . N. J. to e. Tenn. and Fla.", and more recently Glea- 

 son (1952) has given the distribution to be: ". . . N. J.; . . . Va. to 

 Tenn. and N. C." Its rediscovery in Kentucky would be a great find. 



In the second category of "lost" species of Liliaceae in Kentucky 

 are the following species which have never been reported from the 

 State and which, as far as can be determined on the basis of specimens 



