
          spica 3-4 foliis longae Rhachis andetes deprepo subtetiagona alternis cateribus
planioribus marginoscabris spiculae sepiles alternae apprepo lanceolatae
teretinsculae glabrae denuptes setis 6 lineas longae. Plumae lanceolatae breviacutatae
comcexae Terbacea margrim rumbranaceae nervis 5 prominentibus
cortata inferior 3 lin. superior 4 1/3 lin. longa. Rhachila brevis fragiles. Flosculi
imbricati lanceolati dorso conueri 5 lin. longi 3-4 inferiores setigeri supremi
mutici submuticire incompletiores valvules inferior chactacea bicentates
ab inferiori parte terinata 3 nervis apic 5 nervis nervo medio in setans validam
7 lineas longam sicuitates ad angulum rectum patentum humido staten rectam
scabians procurrento. Ludicidae ovario paullo breviores lanceolatae acutae
ciliatae menibranaceae ovarium more generis pyriformi conicum vertice hirto
stigmata flumiose (I have omitted one or two things of less consequence)
Elymus canadensis Lin. var. dursiflorus spica densiore breviori spiculis
omnibus pere ternatis setis longis recurvo patulis foliis angustioribus complicatis
forsan E. villosum festinet St. Louis. Elymus virginicus L.
St. Louis.

Arlary 13 February 1836

My dear Sir,

The above concludes what I provided to copy out
for you in the hope that you might find these notes useful
in the preparation of your flora. you will find many given 
with usually adopted and in the present state of Agrostografley
I scarcely know what to advise. Raspailles system is fully
Palisseth ab Beauvier is good as far as it goes- I do not mean
his disposition of the genera, but merely his flitting of the genera;
many of them however are founded on too slight grounds. Trimius
is excellent but perfectly artificial, so that he adopts only those
genera that he can thrust into a Clavis analytica, and to which
he can give abridged characters- his generic characters are therefore
not natural , and he , as in the case of Andropogon, pops everything
into it and makes it as great a chaos as did Linnaeus-of Kunth
I spoke in one of my former letters: his Enum. Gram. is a terrible
map of confusion: many are the species he puts into the genera with the 
characters of which they do not agree- (Nees v. Essenbeck) does refer the 
principles laid down by Brown-although he makes more trolies
in the order; he has exposed his plan in his Agrost. Brasiliensis
but in the manuscript he has sent me he falters several points.
of course I follow him in the Prod. fl. Peru or. I see
        