56 Proceedings of the Biological Society of Washington. 



length and the acuteness of the apex. The variation in the species with 

 stinging hairs on the calyx and pods is of the same sort. In Woodville's 

 plate the calyx-lobes are especially elongated, however, but this we sus- 

 pect is in part an error of the artist. At least in no specimens examined 

 are they quite so long and slender. 



The leaf differences referred to are perhaps valid. In all the Stizolo- 

 biums the leaflets vary greatly in size dependent on the vigor of the plant 

 or the relative amount of shade. While the apex of the leaves is usually 

 acute or acuminate and mucronate, they may be obtuse and mucronate 

 on the same plant, especially on small leaflets, but in the plant we take 

 to be the same as Wight's they are mostly obtuse. 



The most important character is the pod, the valves of which are said 

 to lack the longitudinal keel or carina. In all the forms studied the keel 

 is present in matured pods though absent in full-sized unripe pods. This 

 fact gives rise to the suspicion that Wight was led to error by not possess- 

 ing fully matured fruit. 



In reference to this and other matter Dr. Otto Stapf writes under date 

 of August 26, 1912, as follows: 



" The Herbarium, 



" Royal Botanic Gardens, Kew, 



"26. 8. 12. 



" The plant from which Wight's figure of Mucuna prurita (text) was 

 prepared is undoubtedly from the Presidency of Madras. We possess 

 the original drawing with the Tamil name written on in Tamil charac- 

 ters; but no locality is stated. As to the discrepancy between the name 

 in the text and the plate Wight himself in a letter to W. Hooker of 5th 

 March 1832 deplores it as regrettable, without, however, explaining how 

 it came about. Very likely the plate was finished and struck off before 

 Wight was quite clear about the distinctiveness of M. prurita from Asia 

 and M. pruriens from America. From the letter referred to it also ap- 

 pears that Wight had originally another name for M. prurita in mind, 

 but fell in with W. Hooker's suggestion, although somewhat reluct- 

 antly. 



" We possess a specimen communicated by Wight to W. Hooker and 

 written up as M. pruriens from 'Madras.' It is the plant figured in 

 Hook. Bot. Misc. 11, (suppl. tab. XIII). There are two other speci- 

 mens also from Wight, one with the Herb. Wight propr.' label and the 

 other from the distribution set; but neither has ripe pods, so that your 

 suspicion that Wight's description of the pods as being ' not at all cari- 

 nated ' was due to the pods being immature, is evidently well founded. 



"As to your question (1) is the name M. prurita to be credited to 

 Hooker or Wight? I am inclined to vote for Wight who is responsible 

 for the publication. What Hooker did was antecedent to the publica- 

 tion and the utmost a sticker for historical justice could claim would be 

 M. prurita Hook, ex Wight. 



"Ad question (2), I would reply, the name M. prurita stands on the 

 Madras plant, as described and figured by Wight. Even if it should 

 turn out that Rumphius's Cacara pruritus is not Mucuna prurita Wight, 

 from Madras, but another Mucuna, the rule of the conservation of the 

 earliest specific epithet could not be invoked as Cacara pruritus is pre- 

 Linnean." 



The actual identity of Wight's plant is important from a nomencla- 

 torial standpoint. There are three subspecies closely resembling his 



