RECORDS OF THE BOTANICAL SURVEY OF INDIA. 
3 44 
surface white, densely covered with a short wooly tomentum and nerve 
traces prominent. Petiole 1 cm. long, tomentose, canaliculate above. 
Corymbs 3-6 crn. broad, 3-5 cm. high, with 8-40 capitula. Pedicels 
*4-1 5 cm. long, strongly tomentose, terete and slightly thickened 
towards the capitula. Capitula campanulate, 8 mm. high, 5 mm. in 
diameter, about 1 5-flow •'red, light-rose coloured. Involucral bracts about 
4-seriate, markedly increising in size towards the flowers, the outer 
ovate apicuiate, 1 mm. Jong, *5 mm. broad, only very faintly and sparsely 
tomentose on the upper margins, the inner tending to linear, acute, 
apicuiate, 5 mm. long, 1 '5 mm. broad, glabrous, all purple coloured 
towards the tip. Receptacle convex, 1*5 mm. in diameter, plainly reticu- 
late. Corolla glabrous, about 8 mm long, of which the linear lobes 
3 mm. Lobes somewhat thickened or folded on the margins with a 
blunt brownish tip (in dry specimen). Stamens epipetalous on the 
middle of the corolla and united throughout the whole length of their 
anthers; filaments short and slender, 1*5 mm. long; anthers twice as 
long as the filaments. Style slender, as long as the corolla, glabrous 
throughout J of its length, slightly thickened and papillose at the 
top and diverging into two linear to awl-shaped, arching, stigmatic, 
papillate lobes. Achaenia ten -ribbed with small amber-coloured glands 
between the ribs, glabrous, narrowly obtfvoid but truncate at both ends, 
2*8 mm. long, *8 mm. broad at the broadest part. Pappus uniseriate, 
plentiful, ascending to erect, 6 mm. long, minutely barbate throughout 
its whole length. 
Making a possible exception of the grasses more attention lias been paid by systcmatists 
to the composites than to any other group of flowering plants, and within the family the 
section Vcrnonieae has been subject to that investigation which their world-wide distribu- 
tion would lead one to infer the group deserved. Nevertheless material still frequently 
accrues, extending the formerly existing limits of the group or filling up gaps which left 
it disjointed. At best it is little better than a heterogenous collection of species, the sectional 
sub-divisious beiug based ou characters at once vague, variable and unreliable. The 
dependence of individual workers, therefore, on the general facies of a specimen is less a 
taxonomic error than a practical necessity. Following Clarke (Comp, lndicae, pp. 6, 7) 
the plant under description would find a place in the section Gymnanthemum, would have, 
on the one baud, the advantage of a place near the well known V. indica Clarke, with 
which it has well marked affinities, on the other, the disadvantage of proximity to V. 
saligna DC, V. divetgens Btli. etc., with which, in the writer’s opinion, it has little con- 
nection. Hooker’s arrangement (F. B. I. Vol. Ill p. 229) batter suits oar plant. It would 
then come near V. Wightiana in the group that author distinguishes by the presence of a 
dense white or buff wooly tomentum. 
Geographical considerations further! support the advisability of so placing V. Jfysoni 
for the few species defined as above all hail from Peninsular India or Ceylon. 
C. B. Clarke (Comp. lndicae) recognised some 41 Vernonias as indigenous to British 
India. When Hooker revised the genus for the F. B. I, three only were added to the 
number. Gradual additions, nearly all -from Southern India have brought the number to 
oyer 50 at present with some material in the Ca lentta herbarium still unsatisfactorily placed 
