
          of the rarest that I could spare. This I could do without infringement
of orders.
I am glad you are pleased with the Prodr. fl. Penns. [Florae Philadelphicae prodromus], and also
with the Cyperaceae by N. v Esenb. [Nees von Esenbeck]. It is more than I now am, for
I have discovered some very silly errors (chiefly of the press). Then in
the Prodr. [Prodromus] at p. 188, no. 584(34) "legumes ... about 3-seeded" ought to be
8-seeded, and at p. 192, No. 600(50), "peduncles ... bearing 2 small flowers"
ought to be 2-5 small flowers. I have now ascertained that 
Lespedeza Indica (p. 231) which I had not seen, is Rhynchosia rufescens
(p. 239), so that Lespedeza is [crossed out: a truly n] [added: not] a peninsular genus. I may
here add that Hedysarum cuneifolium Roth (see p. 198) is a new species
of Taverniera DC [De Candolle], a genus not in the Prodr. I [crossed out: illegible] now possess
a miserable specimen sufficient to enable me to draw up a specific
and corrected generic char. [character]. As to the "Wight's Contributions [to the botany of India]" [crossed out: ought?]
owing to Nees v. Esenbecks most being very ill written, more blunders 
of importance have occurred, and I have added some of my own. Thus 
at p. 75, l. [line] 9, for referta read relata; l. 23, read umbella simplici, spiculis
alternatim subcapitatis incurvatis 20-40 floris [N?]. This is C. [Cyperus] Linnaeanus of
N.E. [Nees von Esenbeck's] [mat.?]. p. 79, no. 23, in the 2d line, read simplici vel composita 3-6 [radiata?], 
and in l. 3d, delete the word recta. p. 84, l. 2, for revensiti, read
recensiti. Same page, no. 40 C. [Cyperus] canescens, ought I think to be brought
into the former section: to that species [Cheed?] 12, t. 55 belongs.  p. 85, n
44. This I have now ascertained to be the true C. verticillatus Roxb [Cyperus verticillatus Roxburgh], [?]
of that no. l. 1. for digitatus read verticillatus. l. 7 for rubris, read scabris; 
l. 8, for digitatus, read verticillatus, for 205 read 206; and before N. ab E. [Nees von Esenbeck] 
insert C. digitatus. Delete the observation upon this species viz. "C. digitatum
Roxburghii esse puto [?]," and also my note at the bottom of the 
page. p. 87 n. 50, the name verticillatus must be changed:
I do not recollect what I have substituted for it in the [vol?] of Beechey's
voyage where I have noticed it. delete also the syn. of Roxburgh, and
also "In Penins. Ind. or. Roxburgh" and insert before "N. ab E." the words
"C. verticillatus"!. delete the whole of my observation viz [In tabula [&c?]] 
same page, no. 51, [crossed out: for] l. 6, for triquetro [trequetro], read explicata. p. 88,l. 4,
for bifido, read trifido; same for no, 54, l. 2, read umbellatis, umbellulis 
e spicis [added: subsimplicibus] 3-4 sessilibus quibus damque pedunculatis patientissimis,
spiculis alternis [&c?]. p. 93, l. 12, after basi conica acuta, [crossed out: read] [added: insert] sensim.
Delete under that plant [the?] Hab. [Habit] "E. Ceylona et" and add at the end of the 
observation [Rheed. mal. 12, [to?] 58 habitu magis ad H. schoenoidem accedit,
at fide patriae huic pertinet. ad H. schoeroidem certi pertinet Wall. [Wallich] 
cat. n. 3404d, et forsan etiam C. Arn. [Arnott]]. p. 94, no. 3  l. 6, for intimis
read infimis, same [?] last 4 lines, I suspect that Nees v. Esenb. here [?]
Cyp. pygmaeus, and not "C. pumilus "which bears no resemblance to Dichostylis.
p. 97, §. 2, l. 1, delete monostachyis , and at l. 3, after aut simplici, insert radiis

        