262 ERICACER (Guthrie & Bolus). [| Hrica 
tube ; anthers from cuneate to oblong, 1—2 lin. long, crested; pore 
2 the length of the cell or more; crests lanceolate to ovate, incised 
or serrulate, about 1+ the length of the cell; style included; stigma 
capitate. Berg. Pl. Cap. 108; Bauer, Exot. Pl. t. 21; Benth. in 
DC. Prodr. vii. 652. EH. eorifolia, Thunb. Diss. Erica, 46, partly ; 
see Rach in Linnea, xxvi. 779, . corifolia, var. spicata, Wendl. 
Eric. Ie. fase. 24, 185, t. 70. E. calycina, Andr. Heathery, t. 8, 
and Col. Heaths, t. 8; Wendl. l.c. fasc. 10, 11; Lodd. Bot. Cab. t. 
594. EH. calycina, var. melastoma, Andr. Heathery, t. 254. H. 
articularis, Curt. Bot. Mag. t. 423. “ E. tunicata, Bartl. in Linnea, 
vii. 686, and E. Alopecias, Tausch in Flora, 1837, 498,” acc. to 
Benth. £. bracteata, Thunb. Diss. Erica,13. E. obvallaris, Salish. 
in Trans. Linn. Soc. vi. 886, and “ E. hyssopifolia, Salisb, l.c. 387, 
and EH. obcordata, Sinclair, Hort. Eric. Wob. 17,” ace. to Benth. 
E. polygaleflora, Klotzsch in Linnea, xii. 537. E. patula, Klotzsch, 
Le. 538. EH. calycanthoides, Klotasch, l.c. 539. E. obtecta, Tausch 
in Flora, 1839, 638. EH. togatoides, Forbes, Hort. Wob. 86. E. 
erectiuscula, Wendl. ex Klotzsch, l.c., 5386. EH. pigra, Soland. 
ex Salish. in Trans. Linn. Soc. vi. 387. 
Coast ReGion: frequent on the plains and mountains from Tulbagh Div. to 
George Div., Thunberg, Burchell, 7661! 7988! Ecklon & Zeyher, 3267! 3268 ! 
38271! MacOwan & Bolus, Herb. Norm. Austr.-Afr., 21! 1804! Guthrie, 
1305! 2282! 2287! 2502! 2794! 8613! 4591! Bolus, 1199! 4172! 5431! 
6742! 6895! 7512! 8494! 9168! Schlechter, 2267! 5414! 7262! 9508! 
10360! Galpin, 3615! 3617! 4941! Wolley Dod, 295! 894! 904! 
A very variable species, which has been unduly divided by the earlier authors, 
and even Bentham observed under E. corifolia: “A variable species, nor is it 
separated by any certain characters from its neighbours, since it is connected by 
intermediate forms with the preceding species [E. bracteata, Thunb.] and with 
the four following [E. polygalaflora, patula, nodifiora, and teretiuscula].”? We 
have examined a very large suite of specimens, and in the result find that 
it runs into three chief forms, which cannot even be separated into varieties, since 
they are all connected by intermediates. These are : 
A. Corolla 33-5 lin. long; floral leaves often larger and sepaloid ; sepals from 
a little shorter than, to a little exceeding the corolla. ‘This form answers broadly 
to E. bracteata, Thunb. : 
B. Corolla mostly 2-3 lin. long; floral leaves like the cauline ; sepals from 4 
of the corolla to equal to it in length. This seems to be FE. corifolia as under- 
stood by Linnzeus and the older authors, and as figured by Bauer, by Andrews, 
Le. t. 8, and by Wendland, J.c. fase. 24, t. 70. 
C. Corolla 14-2 lin. long, its segments sometimes, but not always, obtuse and 
broader than long ; sepals often white or pallid, in length as in B; leaves often 
all opposite, or 3-nate and opposite on the same plant. ‘These answer mostly 
ee E. patula, Klotzsch, and in part to E. polygaleflora, Klotasch, besides 
others. 
The nearest allies to the species, as we understand it, are EH. nodiflora and EB. 
articularis, 
395. E. nodiflora (Klotzsch in Linnwa, xii. 539, not of Salisb.) ; 
erect, 6-12 in. high, entirely glabrous; branches numerous, ascend- 
ing, slender, sometimes filiform ; leaves 3-nate or rarely opposite, 
mostly about as long as the internodes, or occasionally imbricate, 
linear or narrow-oblong, acute, keeled, suleate, 1-2} lin. long ; 
