48 
AMERICAN JOURNAL OF BOTANY 
[Vol. 8 
var. 7. But in the Systema we have "P. spica tereti, involucellis bifloris 
fasciculata-pilosis (applying to pearl millet and not to yellow foxtail), 
seminibus undulato-rugosis. Sp. pi. n. 2. 7" (applying to yellow foxtail 
and not to pearl millet). Instead of "restricting" the name to either species 
Linnaeus adjusts it to both. In the foxtail the spikelets are solitary in each 
fascicle and the fascicles are not pilose. In pearl millet the spikelets are 
usually two to the fascicle and the fascicles are pilose. In the second edition 
of the Species Plantarum, which affords greater space, Linnaeus again cites 
"Fl. zey. 44" (pearl millet only) and gives the original description of pearl 
millet, adding to it that of the fruit of yellow foxtail. 
Now "in the interests of stable nomenclature" what can be done to 
bring order out of this confusion? 
My own judgment would be that: 
1. Panicum alopecuroides be restricted to that of the Species Plantarum, 
1753, the citations discarded, leaving the Chinese specimen as the type 
(= Pennisetiim alopecuroides (L.) Spreng., but not as Sprengel applied 
the name). 
2. Panicum glaucum be restricted to its first diagnosis and description, 
applying clearly to pearl millet, and that the citations, including the 
varieties /? and 7, be discarded. The name glaucum itself applies to 
the bluish tinge of the spike of pearl millet (= Pennisetum glaucum 
(L.) R. Br., but not as Brown applied the name). 
3. Panicum americanum be rejected, since it was based on two unidentifiable 
figures, one evidently drawn from a tracing of the other. 
4. Panicum cynosuroides be rejected. This seems to have been used but 
once, by Scopoli in 1778, and applied to Chaetochloa viridis. 
5. Holcus spicatus be restricted to its first diagnosis in 1759, becoming a 
synonym of Pennisetum glaucum. 
The object in publishing this lengthy analysis is primarily to make avail- 
able for the use of others what has cost much time and study. But a second 
reason is that it furnishes a good example of the "Linnaean concept of 
species," to which botanists who are not systematists sometimes bid us 
return. This is not to find fault with Linnaeus, it is only to show that he 
was human and fallible like the rest of us. His concept of species was not 
"broader" than ours, as is commonly supposed by those who have not 
used his books. He described very closely allied forms, such as Bromus 
purgans and B. ciliatus, or even "split " a single species, as when he described 
Andropogon divaricatum and A. alopecuroides for the commonest Erianthus 
of our eastern states {E. divaricatus (L.) Hitchc). When Linnaeus had a 
plant in hand his descriptions are often vivid impressionist pictures. That 
of Panicum dichotomum "like a little tree, simple below and branching 
above," recalls the autumnal phase of the plant instantly to one who knows it. 
But in the majority of his species there is no description but the brief diag- 
nosis following the name, which is often inadequate for identification. 
