122 PENNELL: STUDIES IN THE AGALINANAE 
series into a number of definitely restricted New World and Old 
World genera. The Old World genera differ from the New in 
more points than were realized at that time, so may be definitely 
dismissed from the present discussion. 
After Linnaeus’ time the first new names proposed were in 
1788 Walter’s two genera both named Anonymos. These were 
well characterized; the one might be typified by Gerardia purpurea 
L., the other by a new species Anonymos cassioides Walt. Of 
course the name—or confession of the lack of a name—Anonymos, 
has no value in nomenclature. 
In 1791 J. F. Gmelin, reviewing Walter’s work, returned his 
first Anonymos to Gerardia, but maintained his second Anonymos 
as a new genus Afzelia. Anonymos cassioides Walt. became 
Afzelia cassioides (Walt.) J. F. Gmel. This genus also was by 
later authors returned to Gerardia, and when in 1814 Pursh became 
convinced of its generic distinctness, finding the name Afzelia 
meantime applied to a genus of the Caesalpiniaceae, he renamed 
the genus Seymeria. Afzelia J. F. Gmel. was restored by Kuntze 
in 1891. The genus to which this name is applied is a definite, 
natural group of Mexico and the southern coastal plain region 
of the United States. 
The next generic description in this group is in 1794 that of 
Virgularia described by Ruiz and Pavon from Peru, and based 
upon their V. lanceolata, the specific description of which did 
not appear till 1798. Though loath with limited material to 
enter upon any discussion of the South American species of this 
group, it is necessary to attempt to decide whether this genus 
can be distinguished from those later proposed to include our 
North American species. 
In the differential characters pointed out by Ruiz and Pavon 
I find, I confess, little that is convincing. The description of a 
bifid stigma is surely remarkable, but as no later observer has 
recorded such a structure in this or in any other South American 
species of this group, there is doubtless here some error. Also the 
characters depended upon by Martius (1829) have been shown by 
Bentham (1835) to be untrustworthy. Since 1835 the name 
Virgularia has been considered a synonym of Gerardia L. Yet 
an inspection of the species of Virgularia will, I think, show us 
