38 
collection of 1828 which, as already pointed out, cannot be identified with 
either the description or the major part of the plate of T. violaceum in Flora 
Danica. 
Lange’s conception of violaceum also becomes abundantly clear from 
the fact that specimens which through the courtesy of Dr. Morten Porsild, 
the writer has from Ikertbq near Holsteinborg, 1 latitude 66° 58', West 
Greenland, and which are cited in Consp. F. 1. Groenl. under A, violaceum, 
are typical A. latiglume (Scribn. & Sm.) Rydb. and consequently do not 
agree with A. violaceum of Flora Danica. 
In the circumstances A. violaceum Lange must be considered a later 
homonym and should be rejected, according to Art. 51 bis, adopted at the 
Fifth International Botanical Congress. 
Koch’s T. biflorum Brign. j3 Hornemanni, based on T .violaceum Hornem., 
applies to a plant with purple spike and 3-nerved empty glumes. Accord- 
ing to Ascherson and Graebner (1, page 654), Koch’s Hornemanni has 
the empty glumes wide and abruptly contracted toward the awned apex. 
If so, the name Hornemanni applies to A. violaceum Lange, but not to T. 
violaceum Hornem. In other words, it refers to forms of A. latiglume 
(Scribn. & Sm.) Rydb. Pease and Moore, in the key to the varieties 
and forms of A. caninum (13, page 69), separate var. latiglume and 
var. Hornemanni , on the presence or absence of pubescence on the lemma, 
the former having “glumes with a fine pubescence,” and the latter having 
“glumes minutely ciliate on the margins, not pubescent.” They cite, 
under var. Hornemanni , among others, one specimen in the National Herb- 
arium of Canada, No. 26026, from Chilliwack valley, British Columbia. 
This is A. latiglume w T ith glabrous flowering glumes. 
In the preceding it has been pointed out that J. Vahl’s collection of 
1828 represents A. latiglume and that the collection from 1829 with green 
spikes represents another species. This green-spiked species was separated 
by Lange from his A. violaceum as /3 virescens. Through the courtesy of 
Dr. Porsild, the w r riter has before him a specimen from Kugssuaq in Taser- 
miut-fiord 2 60° 13' north, collected by J, Vahl, August, 1829, w r hich has 
belonged to the herbarium of Lange and which is cited by him under A. 
violaceuyn /3 virescens . This, J. Vahl, according to information kindly 
supplied by Dr. Porsild, later identified w r ith “ Tr . biflorum.” T. biflorum- 
A. biflorum (Brignoli) R. & S. — however, has nothing to do with the 
Greenland plant; neither does it belong to the North American continent. 
It is, as the writer has satisfied himself from examination of authentic 
specimens, kindly loaned by Dr. G. Samuelsson, Stockholm, a nearly awnless 
variety of the true A. caninum of Europe, A. caninum var. biflorum (Brign.) 
Richter. 
The late Tycho Vestergren considered A. violaceum (Horn.) Lge. j3 
virescens Lge. a distinct species. On the folder originally marked 0 vir- 
escens in the herbarium of the Botanical Museum, Copenhagen, he has 
made the following note (unpublished): “With the exception of the upper- 
most specimen, all is A. ?nutabile Drob., which appears to occur on the 
southern tip of Greenland (about latitude 60-61), and is the same as Lange’s 
/3 virescens .” A. mutabile, however, is apparently not identical with the 
Greenland plant. It is a species of northern Fenno-Scandia, northern 
Ukertok-Fjord v. Holstenborg, as Lange's writing reads. 
iKorsoak i Tasermiut-Fjord, according to Lange’s spelling. 
