126 
roup ‘‘ Panicatrix’’ in allusion toe the habit of the panicle 
which is that of a typical Panicum. But whilst Setaria longiseta 
is no doubt a Setarza, it is not the “‘ type’’ of Setarta. Beauvois 
himself (l.c.) in the heading over the description of the genus 
quotes “‘ Setaria Ess. d’Agrost.’’ If we turn to his Essai d’une 
nouvelle Agrostographie, we find (1) on p. 51 the description of 
the genus ‘* Setaria nob.,’’ and a list of species to be transferred 
from Panicum to Setaria; then (2) on p. 178 (index) an enumera- 
tion of the new names under Setaria; (3) on p. 9 of the “‘ Explica- 
tion ’’ the explanation of the figure of Setaria viridis which is 
intended to illustrate the genus; and lastly (4) on tab. 13 under 
fig. 3 a good picture of a spike of Setaria viridis with equally 
excellent analyses. Several of the combinations on p. 178 are 
queried; of those without a sign of interrogation all but two are 
recognised members of the Hu-Setarta group, namely germanica, 
glauca, italica, setosa, verticillata and viridis, the latter of whic 
was, as already stated, chosen by Beauvois to illustrate the genus. 
The two doubtful species are S. purpurea and S. sericea, both 
taken over from Panicum. Of these S. purpurea is probably a 
true Setaria, whilst S. sericea is evidently Pennisetum setosum. 
However that may be, there can be no doubt whatever that 
Beauvois meant his Setaria to include the species of the type of 
Setaria viridis, and this will have to be considered as the founda- 
tion of the genus. Scribner’s mistake in taking Beauvois’ Setaria 
longiseta, the only species mentioned in his Flore d’Oware et de 
Benin, as the ‘‘ type ’’ of the genus arose probably from the mis- 
leading date on the title-page of the second volume of that work. 
This is given as 1807, that of the Essai d’une nouvelle Agrosto- 
graphie being 1812. The Flore d’Oware et de Benin was actually 
published in parts, those of volume ii. being spread over fourteen 
years (1808-1821). The dates of part 11 (pages 1-12, tt. 61-66), 
with which vol. ii. starts is 1808, of 12 and 13 (pages 13-32, tt. 67- 
78) 1810 (see Hallier in Jahrb. Hamburg. Wissensch. Anst. 
xvii. 67), whilst it appears from a notice in Flora v. part i. (1882) 
Beibl. 1. p. 4, that only seventeen parts had been published by 
1818. According to Hiern (in Journ. Bot. 1898, p. 495) part 17 
contained tt. 97-112. The date of the publication of S. longiseta 
is therefore six years later than that of the Essai d’une nouvelle 
Agrostographie. 
