RICH O D ESM A INRICUM a'nd T. AMP IFXICAULF 
349 
The specimens of this form, so far as the material in the Herbarium 
of the Agricultural College, Poona, is concerned, are mainly from the wet 
tract of the Konkan. It may therefore be purely a climatic variety. 
On the other hand it may be a hybrid between the two species. This 
is what I take to be the T. amplexicaule of Roth, as will be explained 
below, and should for the present be regarded as a variety of T. indicum. 
As regards synonymy there is no question about Robert Brown's 
plant. Brown in his Prodromus Florae Novae Holiandiae p. 496 
merely transferred Linnaeus' Borrago tndica to his 5wn new geuus 
Trichodesma. The description of Brown's T. indicum in DeCandolle'' s 
Prodromus, Part X, p. 172, is quite clearly the description of the plant 
described in the left hand of the two parallel columns above. In the 
Sub-sectional heading DeCandolle says — “ Anther a exserta, apice 
eiongato contorts 
When we come, however, to Roth's description of his T. amplcxi- 
caule in his Nov. Plant. Sp. p. 104, we find that it is not the descrip- 
tion of the plant of the right hand of the two parallel columns, but 
is the description of the form of T. indicum which I have referred to 
as var. amplexicaule , since Roth describes the andrcccium thus 
w Anther a erectec , subnlata , in fasciculum pyramidal em apice spiralcm 
conniventesM This cannot possibly refer to anything but T. indicum , 
while the rest of the description is of the leaves and habit of my var. 
amplexicaule . 
On the other hand the description of T. amplexicaule in DeCandolle's 
Prodromus, (l.c.) is equally clearly the plant described in the right 
hand of the two parallel columns above. In his heading to the sub- 
section which contains only the one species DeCandolle says : — 
“ Anther a inclusa, apice non contorts. Corolla lobi breves , mucro- 
nnlatiJ ' And in his description' he says : — “ Stamina inclusa , in conum 
villosum conniventia , apice non contorta y villis erectis apicem gla - 
brum aequantibusR This is quite clear. DeCandolle was however 
wrong in thinking that the aestivation was quincuncial. It is contorted, 
though the rest of the details of the flower buds are quite different 
from those of the other - species. DeCandolle had only one specimen 
of Roux to go by, and he sa,ys—“ et aestiv. corolla verisim. diversa 3 
quodequidem mirandum , sed ex 2 floribus solum observare potuid* 
The key to the three plants is then : — 
1. Anthers exerted, c ombined into a complete cone ; back of 
the connectives densely tomentose with short, crisped, 
yellowish-white hairs ; tips elongate, glabrous, spirally 
contorted. 
(a) Leaves crowded, very narrow, villous, with few or 
no tubercle -based hairs . .. . , . . . T. indicum R. R r . 
(type). 
