1897.] D. Prain — Some additional Leguminosse. 369 



water better than many others. But it is for its tall and slender stems Which some- 

 times reach 20 feet in height without being more than '5 in. thick at the base, and 

 which are always hard, never soft and pith-like as in 8. paludosa, that the plant is 

 mainly cultivated ; these long lithe stems are nsed as the wattles of which are 

 constructed the walls of the houses wherein Piper Betle is grown. 



5. Sesbania uliginosa Sweet Hort. Brit. 129 ; diffuse, unarmed, 

 racemes short few-fld., pods erect subtorulose, not twisted. Aeschyno- 

 mene uliginosa Roxb. Hort. Beng. 56, Flor. hid. iii. 334. S. paludosa 

 Jacq. in DG. Prodr. ii. 265 not Aeschynomene paludosa Roxb. 

 Bengal ; in swamps. 



This species Dr. Roxburgh compares with the South Indian 8. procumbens, and 

 Wight and Arnott would endorse this comparison. What these authors mean 

 precisely when they say that Roxburgh's A. uliginosa is not the 8. uliginosa of 

 " authors," is hardly clear, for there would appear to be only one published 

 S. uliginosa, that of Sweet, which is founded, Sweet indicates, on the Aeschynomene 

 uliginosa of the Hortus Bengalensis. It is just possible that the 8. uliginosa 

 referred to by Wight is S. paludosa (Aeschynomene paludosa Roxb.) 



6. Sesbania aculeata Pers. Synops. ii. 3J6 (excl. citations Plunkenet 

 and Rheede) ; DO. Prodr. ii. 265; Bak. in Flor. Brit. Iud. ii. 114 (excl. 



all the varieties) . 



A weed of rice fields and swamps throughout India. Two moi'e or less distinct 

 varieties are recognisable. They differ, however, only in habit, the flowers and fruits 

 of the two are identical and intermediates are numerous. 



1. typica; stems reddish, rather densely sprinkled with minute prickles. 

 Aeschynomene spinulosa Roxb. Flor. Ind. iii. 333. S. aculeata W. 8f A. Prodr. 215 

 {excl. syn. A. cannabina Roxb. and A. bispinosa Jacq.). 



2. Var. elatior ; stems green, sparsely prickly, taller, lax and slender. Aeschy- 

 nomene bispinosa Jacq. Ic. Rar. iii. t. 564. Coronilla cochinchinensis Lour. Flor. 

 Cochinchin. ii. 552. S. cochinchinensis DG. Prodr. ii. 266. S. cannabina W. Sf A. 

 Prodr. i. 215 not Aeschynomene cannabina Roxb. 



Wight and Arnott refer Roxburgh's Aeschynomene cannabina, the Dhunchi plant, 

 to 8. aculeata. This is so obviously wrong that the only conclusion to be formed is 

 that they never had an opportunity of examining a Dhunchi plant. And that they 

 are right in regarding their 8. cannabina (which the writer cannot separate specifi- 

 cally from their 8. aculeata) as the plant that Retzius named Aeschynomene cannabina 

 and that Willdenow named Coronilla cannabina, is highly improbable. The descrip- 

 tion given by Retzius really only fits well, among Indian Sesbanias, Wight and 

 Arnott's own S. procumbens. It is merely the fact that Retzius has said, on Koenig's 

 authority, that the plant to which he refers is the fibre-yielding species (and 

 therefore the Dhunchi), which has led Roxburgh, whom the writer is quite willing 

 to follow, to apply the name " cannabina " to the Dhunchi plant. 



7. Sesbania procumbens W. 8f A. Prodr. i. 215; Bah. in Flor. Brit. 

 Ind. ii. 115. Aeschynomene procumbens Roxb. Flor. Ind. iii. 337. 



As already mentioned this is the species which best fits the description given bv 

 Retzius of Aeschynomene cannabina; the reasons that have led the writer to adopt 

 the Roxburghian interpretation have been stated. 



