37 
As Koch’s var. Hornemanni was based on Triticum violaceum Horn., it 
is obvious that, in order to come to an understanding of what var. Horne- 
manni really is, it is primarily necessary to investigate the identity of 
Hornemann’s T. violaceum. There is no material designated as the type 
of this species, but it is explicitly stated, in the explanation of the plate in 
Flora Danica, that it is based on material collected by J. Vahl in southern 
Greenland. Now, there are, in the herbarium of the Botanical Museum, 
Copenhagen, Denmark, two collections of Agropyron by J. Vahl, from 
southern Greenland, made prior to 1832, the year T. violaceum was pub- 
lished, one from 1828 and another from 1829. 
The 1828 collection, from latitude 60° 54', consists of three sheets 
with a total of five specimens. In all the specimens, which have more or 
less purplish spikes, the empty glumes are broadly oblanceolate or even 
obovate, abruptly contracted toward the apex, 3-nerved or occasionally 
4-nerved, and with a broad, hyaline margin. All five specimens have the 
lemma more or less pubescent. None of them agrees with the description 
of T. violaceum which says that the empty glumes are 5-nerved. Neither 
does any of them agree with the figure of the spike in Flora Danica. One 
of them, however, agrees perfectly with the detail of the figure of the 
flower, having a lemma that is glabrous in the upper part and pubescent at 
the base, exactly as pictured. All the specimens of the 1828 collection 
belong to A. latiglume (Scribn. & Sm.) Rydb. 
The 1829 collection, from latitude 60° 13', consists of two sheets with 
one specimen each. Both have green spikes. The empty glumes are 
lanceolate, gradually contracted toward the apex, 4-6-neived, and without 
a conspicuous, broad, hyaline margin. The lemma is glabrous throughout 
and not at all pubescent at the base, as figured in Flora Danica. These 
specimens agree, as to the general appearance of the spike, with the figure 
in Flora Danica, but the spike is green and not purple, as figured. The 
empty glumes also agree with the figure as well as with the description. 
This collection represents a species different from A. latiglume. 
Considering the characteristics of the two collections just mentioned, 
in comparison with the figure of T. violaceum, one cannot but gain the 
impression that the latter is a composite one, the lemma, with its peculiar 
distribution of pubescence, and the colour of the spike, being taken from 
the 1828 collection, the general appearance of the spike, including the shape 
of the empty glumes, from the 1829 collection. This is, indeed, most 
probably the case as Hornemann, who had both collections before him, 
did not realize that they represented two different species. If the assumpt- 
ion is correct, the name T. violaceum Horn, should as a nomen confusum be 
rejected in accordance with Art. 51, sexies, of the International Rules of 
Nomenclature, as approved by the Fifth International Botanical Congress 
at Cambridge, England, 1930. 
Concerning Agropyron violaceum Lange (7, page 155), it should 
be recalled that this is based on Hornemann’s Triticum violaceum. Lange, 
however, modified Hornemann’s original description, applying the name to 
a plant “palea inf. abrupte (nec sensim) in aristam excurrente,” thus 
making the description of his violaceum 1 plainly applicable to J. Vahl’s 
l Lange attributes, through modesty, the transfer from Triticum to Agropyron to Rink, though it was actually 
made by himself (note from Dr. Porsild). 
