REPORT FOR I904. 
37 
Kceleria valesiaca, Gaud. In grassy turf on carboniferous lime- 
stone rocks, in full exposure at Uphill, N. Somerset, i6th Oct. 1904. 
The history of this grass is peculiar. In preparing a memoir of 
Dillenius and an account of his Herbarium, which is now in the 
press, I came across a small parcel containing plants, among them 
being an unlocalised grass, which I thought was a foreign form 
of Kceleria. Subsequently I found a MS. note which almost cer- 
tainly referred to it. Dillenius there says that he found it at Uphill 
and on Brent [Brean] Down plentifully, and he calls it ‘ Spartiu?n 
montanum radice bulbosa et fufigosa.' He identifies it with the grass 
which A. Scheuchzer in the Agrostographia of 1716, p. 169, calls 
‘ Gra 7 tie 7 i valesianum tefiuifolium, patiicula spicala, viridi argentea, 
splendcnte.'' This identification by Dillenius was correct, and it 
testifies to his great critical acumen in recognising it from the 
diagnosis in that work. In the celebrated description which Dil- 
lenius gives of his journey into Wales and the Western country in 
1726, which is printed in the Richardson Correspondence number 
100, Dillenius refers to finding this grass, the identity of which at 
that time he had not ascertained. Dillenius in his description of 
the grass, says when he gathered it in the middle of July, on 
Brent [Brean] Down and Uphill [Somersetshire], ‘ the plant had 
either shed its seeds or had produced none ; ’ and when I first re- 
cognised this plant last October, I thought there was only a remote 
chance of discovering it so late in the season. However, I went 
down to Weston, and within ten minutes of getting out of the cab 
at Uphill, succeeded in finding it growing on the sides of that pic- 
turesque hill. The grass is instantly recognisable from its curious 
rootstock which is covered with a fibrous mass, consisting of the 
persistent strands of the old leaves. The texture of the leaves 
themselves is also different, and the panicles are continuous and 
not more or less interrupted as in K. cristata. Plants with the 
dried flowering stems were also found in sheltered situations, and 
it grew in considerable quantity, especially on the exposed lime- 
stone ledges, which it appeared to prefer. Koderia valesiaca, 
Gaud., has considerable variation; for instance, Willkomm and 
Lange in the Prod. FI. Hispan. vol. i. p. 75, give three varieties, 
namely, var. glabra, Godr., which they consider to be K. valesiaca. 
Gaud. ; var. ciliata, Godr. ; the K . setacea of De Candolle ; and 
var. pubescens, the Aira vallesiaca, Bert. ; but they adopt the 
name K. setacea, Pers., for the aggregate species. As the char- 
acters by which these varieties are separated depend upon the 
inflorescence, I defer the identification of the segregate, and the 
description of our British jDlant until I have gathered it in flower. Al- 
though the authors of the Flora Hispanica adopt the name K. setacea, 
Pers., as a matter of fact, Persoon did not use it. In vol. i. p. 97 
of the Synopsis, Persoon gives the name K. tuberosa, the type of 
which is identified with Aira cristata of Smith, which is our K. 
