14 
or spur, sometimes retuse or slightly emar- 
ginate at the extremity, and the capsule is 
remarkably compressed and truncate or 
emarginate at the top. 
The genus was first proposed by Ven- 
tenat, (Jard. Malm. t. 41.), for a species 
which I do not recognize in any of the 
specimens before me, unless it be a culti- 
vated variety of N. linearis of the same 
author. The latter species is probably the 
Antirrhinum capense,'Thunb. Ventenat's 
third species, N. chamedrifolia, or Antir- 
rhinum macrocarpum, Ait., is evidently 
the same also as Ant. scabrum, Thun 
Persoon added to the genus the Ant. bt- 
corne, Linn. and I now refer to it Ant. 
barbatum, Linn., (Ant. pinnatum, Linn.), 
and eighteen new species. Ant. fruticans, 
Thunb., must also be a Nemesia, but not 
any species that I have seen. Of the re- 
maining Cape Antirrhina of authors, the 
A. aphyllum, Linn., is evidently an Utri- 
cularia ; A. patens, Thunb., “floribus ter- 
minalibus solitariis," is very doubtful, most 
probably not a Nemesia. So also the A. 
dentatum, Poir., with alternate leaves, and 
coming from a country far removed from 
the range of Nemesia, as at present known. 
Two low creeping plants, the one from 
the eastern districts of the Cape Colony, 
the other from Madagascar, with the habit 
of the section Cymbalaria of Linaria, the 
corolla nearly that of Nemesia or Linaria, 
but with a capsule different from either, 
have induced me to propose under the 
name of Diclis, a new genus intermediate 
e 
SYNOPSIS OF THE HEMIMERIDE/E, A TRIBE OF SCROPHULARIACEX. 
lonia, Humb. et Bonpl. Pl. ZEquin. 2. 92, 
which agrees with the old character of 
Hemimeris, although it be, in fact; distin- 
guished as well by some differences in the 
form of the corolla, as by the capsule, 
which is loculicidal with entire valves, in- 
stead of being septicidal with bifid valves. 
The genus now contains four already de- 
scribed species and one new one, besides 
the A. campestris, which has been consi- 
dered by Nees von Esenbeck as a separate 
genus under the name of Thylacantha, on 
account of its indehiscent fruit. The ma- 
terials before me are not sufficient to ena- 
ble me to form myself any opinion on the 
validity of this division. 
The whole of the above genera may be 
considered as forming a small tribe, allied 
to Antirrhinee on the one hand, and to 
Verbascee on the other. They are distin- 
guished from the former by the valvular 
dehiscence of the capsule, and in many 
cases by the double spur, pouch, or protu- 
berance, which is in Antirrhinee always 
simple, and from Verbascee and all other 
Scrophulariacee by the presence of the 
spur, pouch, or protuberance, whether sim- 
ple or double.t 
I now proceed to give the characters of 
the genera and species, beginning with that 
of the whole tribe. 
HEMIMERIDE E. 
Calyx 5-fidus vel 5-sepalus. Corolla 
tubo brevissimo, limbo explanato subrotato 
bilabiato vel personato, late 4—5-lobo basi 
between the two last-named. The capsule, 1. 9 calcarato vel saccato vel bifossulato. 
however, being two-valved (whence the s¥,Jys simplex, stigmate parvo subcapitato. 
name), indicates a greater affinity to Neme- Capsula bivalvis, valvulis integris bifidis 
sia, and places it in the tribe of Hemzme- vel bipartitis. 
2. 
Having thus traced the connecting chain 
tween Hemimeris and the Antirrhinee, 
through Diascia, Nemesia, and Dichs, We cota, Capsula indehiscens. 
may in the same manner connect it with o JAaxuxraan Locals mM 
Alonsoa and other rotate Verbascee, Copeula loculicide bi 
through another South American genus, l 
which has the remarkable concavities of 
the corolla above referred to, besides a 2 Collinsia and some Digitalee have sometimes a 
bebé i Senes TM f slight protuberance on the upper side of the tube, but 
abit in some — eT TE —— y that o very different from those of Hemimeridee, which are 
the racemose Diascie. This is the Ange- always at the base of the limb. 
CoNSPECTUS OF THE GENERA. 
1. THYLACANTHA. Corolla basi sac- 
EMIMERIS. Corolla basi bifossu- 
