1922] The Eastern Shrubby Species of Robinia 177 



the upper surface and white satiny canescent on the lower surface as 

 they unfold and permanently more or less appressed pubescent. Ra- 

 cemes 6-9 cm. long, 3-7 flowered; flowers large 20-23 mm. long, rose- 

 purple or purple and white with a large pale spot on the vexellum; 

 the calyx 12-15 mm. long, the lobes, gradually acuminate and much 

 longer than the tube, purplish or whitish silky canescent, as well as 

 the pedicels or often with short weak bristles intermixed. The fruit 

 mostly solitary at the ends of the short pubescent and often some- 

 what hispid peduncles is dark brown, 4.5-6.5 cm. long, tipped by the 

 stout slightly curved style, the slightly margined body about 1 cm., 

 wide and rough bristly where distended by tlie seeds, but the intervals 

 often much constricted and nearly free from bristles. 



This plant was first collected by me on the south slopes of Grand- 

 father Mountain in 1900 and a seed-bearing plant which attained a 

 height of more than 4.5 m. was cultivated for many years. An in- 

 fertile plant less hispid and more pubescent, the stock of which was 

 collected by me on Grandmother Mountain, in 1915, has been in cul- 

 tivation since that date and numerous plants from it distributed. 

 This species is evidently closely related to R. hispida L., but is readily 

 separated by its larger size, by its pubescence, paler flowers and the 

 fact that it is less hispid. Type, W. W. A., Grandfather Mountain, 

 N. C, July, 1900. 



Many plants of this species, of R. hispida and of R. longiloha, 

 although having perfect flowers do not set fruit. Meehan apparently 

 first called attention to this in the case of R. hispida in cultivation. 

 Large patches of R. longiloha have been examined, all evidently from 

 the same stock by vegetative propagation without signs of fruit — 

 and it is possible that certain patches may not fruit, just as the plant 

 of R. grandiflora which I now have in cultivation has never set fruit 

 — but this I have attributed possibly to the lack of cross fertilization. 

 T have recently obtained fruit of what seems to be typical hispida and 

 have a plant from the seed now growing. 



Forest Service, 

 Washington, D. C. 



