220 THE JOURNAL OF BOTANY 
— (§ Ansexu1a) tanalensis, sp. nov. Caule debili deorsum 
cante ramulis sepe aliquanto tortuosis cito glabris, foliis par- 
vilis distincte petiolatis oaeey lanceolatis utrinque obtusis sub- 
coriaceis glabris, spicis m folia fociaribn 3-8-floris, bradiela 
bracteolis ue abbrevi ete 0 eats obtuse acutis TT ciliolatis, 
euleutibnd. < corolle tubo ealy cem excedente’ oles levied dilatato 
bio postico breviter hikes antici lobo intermedio quam laterales 
fere duplo latiore, antheris subinclusis, capsula sursum angustata 
calycem duplo excedente glabra. 
ab. Ambohimitombo Forest, Tanala ; Forsyth-Major, 414. 
Foliorum limbus 1°3-2°0 ¢ m. long. 0-3-0" 6 cm. lat., cystolithis 
copiose indutus; costa media pubes prominens, pe secundariz 
utrinque - valde arcuate, sepe 8 perspicue ; petioli — 
0°3-0°4 cm. long. Spice modice 2 
bractoolagne circa O°1 cm. long. Calyx 0:4 cm. long. Corolle 
tubus 0:6 cm. long., basi 0°15 cm. superne 0: 95 on em. diam. ; labium 
posticum Sg te in toto 0°3 em. long., lobi oblongi, 
obtusissimi, 0-12 cm. long.; labii antici lobi laterales 0-2 x 0°12 cm., 
lobus intermedius 0-22 cm. lat. Antherarum pone superior basi 
acutus 0°18 em. long., loc. inferior calcaratus 0°15 cm. long. 
Ovarium glabrum, 0-1 cm., stylus puberulus, 0- vi em. long. Cap- 
sula 0°75 cm. long. 
Differs from J. delicatula Elliot in the ¢ dificrently eel leaves 
of firmer consistence, the larger calyx an rola, 
Hyprorstes spscenpens Nees in DC. . i in , is a species 
hitherto supposed unrepresented in this Beary so far as the type 
is concerned. Some fragments from specimens in the British 
Museum, collected by Hilsenberg and Bojer, which seemed to answer 
Nees’s description, were recently sent to M. Casimir de Candolle, 
with the request that he would compare them with the type in the 
De Candolle Herbarium at Geneva. This request was very kindly 
complied with, the result being that the specimens agree with the 
said type except in having 2-flowered involueres, while the type has 
th owered. The involucres which I examined, however, had 
bu va sing gle flower each, and the endeavour to find one with two 
flowers failed ; it would oe therefore, that M. de Candolle 
happened to o an abnormal involu 
at wt be: a second small sterile flower, which fear 
one would im " Selddaobally grow up into a fertile I 
eebitino, thenatores w consider that in these specimens the Vesa 
the true H. adscendens Nees, and as they seem os identical 
with the plant Pogo! desribed under the name 
minthoides by Mr. Baker (Journ. Linn. Soc. (Bot.) x. 222), the 
ame must disappear 
Specimens of oe at the Museum are Baron, 863 and 4153, and 
Hildebrandt, 3487, 
The H. sired of Herb. Kew., quite a different = I rears 
ass now spécian, of Wii? saa 2 en below. 
