174 BALFOUR—NEW SPECIES OF PRIMULA. 
beneath, crinkled and lettuce-like in texture, deeply and sharply 
erose-dentate, clothed on the upper surface in a very minute 
and dense coating of glands, microscopically velvety on inspec- 
tion. Scape + = the leaves, stout, lengthening and stoutening 
in fruit; bracts very narrow and acute, about 6 mm. long. 
Pedicels erect and graceful, about ? of an inch, stiffening but 
not lengthening in fruit, and thickening into a straight, fat 
trumpet in a way suggestive of the Davidi group. Flowers 
from 2-8 in a wide lax umbel; calyx-lobes not = 4 the calyx, 
broadly oval-pointed, sometimes faintly cut into two or three 
teeth, at first standing only slightly away from the tube, but 
then widely apart, enlarging, broadening, and thickening re- 
markably in fruit—as in And. maxima. Corolla-tube white and 
straight, about twice the length of the calyx, widening suddenly 
at the throat, with a slight constriction above in the mouth, 
where, at the base of each lobe, it ends in a bilobed white ray, 
with a greenish tinge below. Flowers 3 inch or more, flat: 
lobes + broadly heart-shaped, lilac-rose, emarginate, scentless. 
Capsule a flattened orb, from which stand out and above 
the fattened calyx lobes: seed borne on a_ cushion- 
disk (?).”’ 
“General in the situations indicated, light woodland, cool 
gorges, etc., in the forest zone of the Siku district, Siku, Satanee, 
Chago, Ga-hoba. Sometimes in limestone loose silt in deep 
cafions : very prolific on rotten fallen tree-trunks, 7000—g000 ft., 
flowering in April and May.” 
The plant is, as Mr. Farrer suggests, one of the Davidi 
Section. We know little of this section. It includes some 
beautiful species, and none are now in cultivation. A plant 
was obtained by Veitch about 1906 from seed collected by 
Wilson, and was introduced as P. ovalifolia, Franch., one 
of this section. It was figured in the Gardeners’ Chronicle. 
It died out in cultivation. I am not at all sure that Veitch’s 
plant was true P. ovalifolia, Franch. More than one plant 
_ appears under the name in herbaria, as happens also in the 
case of P. Davidi, Franch. I hope ere long to be able to 
study the species of the alliance more fully. Meanwhile I may 
recall that neither P. Davidi nor P. ovalifolia has near 
relationship with the suffruticose species P. bracteata, Franch., 
P. bullata, Franch., and P. Henrici, Franch., with which Pax 
unites them in his section Bullatae.* The Section Davidi in 
which I place them has close connection with Sections Petiolaris 
and Sonchifolia—the three showing a characteristic enlargement 
of the scape and pedicels in fruit ripening, and for the fruit itself 
See Primulas of the Bullate Section in Trans. Bot. Soc. Edin. xxvi (1912-13), 
188. 
