ENUMERATION OF THE SPECIES 147 



acutish and acuminulate at the apex and densely grayish pubescent. Leaves 

 elliptic to obovate or oblong-obovate, 3 to 9 cm. long and 1.5 to 3.5 cm. broad, 

 acute or obtusish and gland-tipped at apex, cuneate at base, setosely ciliate, finely 

 pubescent on both sides, more densely beneath, particularly on the veins, rarely 

 glabrescent above at maturity; petioles 3 to 8 mm. long, pubescent and glandular 

 and usually sparingly strigose. Flowers appearing shortly before or with the 

 leaves, end of March or in April in 8- to 15-flowered umbel-like racemes; pedicels 

 0.5 to 1 cm. long, pubescent and glandular-setose; sepals broadly ovate to oblong, 

 unequal, 1 to 2.5 mm. long, glandular-ciliate, pubescent outside; corolla yellow and 

 orange, the tube usually more or less purplish or with 5 longitudinal purplish 

 stripes, slightly fragrant, the tube cylindric, abruptly dilated at the apex, 1.5 to 

 2 cm. long, finely pubescent and stipitate-glandular, the lobes ovate 1 to 1.5 cm. 

 long and 0.6 to 1 cm. broad, short-acuminate with recurved point, finely pubes- 

 cent outside; stamens nearly three times as long as tube, the filaments pubescent 

 below the middle, the anthers ochraceous, 2 to 3 mm. long; style slightly exceeding 

 the stamens, 5 to 6 cm. long, short-pilose near the base; ovary covered with long 

 whitish strigose, partly gland-tipped hairs. Capsule oblong-cylindric, slightly nar- 

 rowed toward the apex, 1.5 to 2.5 cm. long, covered with a thin fine pubescence 

 interspersed with long strigose partly gland-tipped hairs. 



This species has a very restricted distribution; it has been found only in Gads- 

 den County, in northern Florida, near River Junction and Chattahoochee, close 

 to the Georgia state line and about ten miles farther south at Aspalaga near Rock 

 Bluff, Liberty County (Biltmore Herb. 6088). It was apparently first discovered 

 by Dr. Chapman some time before 1865, as he refers to it in his Flora under R. 

 nudiflorum as a color variation of that species; there is also a specimen collected 

 by him in the National Herbarium without date and locality with the note "flowers 

 orange." The species was again collected in 1880 by A. W. Curtiss and distributed 

 as R. nudiflorum var. luteum, but that name was never published and this Azalea 

 was not recognized as distinct until described as a species by Dr. Small in 1913. 



Rhododendron austrinum is so closely related to R. canescens that one might be 

 induced to consider it only a color form of that species, but aside from the fact 

 that the yellow color is a disconnected deviation from the pink color tones of 

 R. canescens, R. austrinum differs further in the glandular ovary and capsule, 

 entirely destitute of glands in R. canescens, and in the copious glandular pubescence 

 of the calyx and pedicels which are usually, though not always, glandless in R. 

 canescens. In their general appearance, however, I could see no difference except 

 the color between the two species, when I saw them growing side by side in Mr. 

 H. H. Hume's garden in Jacksonville, who had brought the plants of both species 

 from River Junction the year before. 



The species was first introduced into cultivation through the Arnold Arboretum, 

 where plants were raised from seed collected in 1914 by Mr. T. G. Harbison near 

 River Junction. It cannot be expected to thrive in northern gardens, but for 

 southern gardens it may be desirable on account of its yellow flowers. 



Rhododendron atlanticum Rehder, comb. nov. 



Azalea nudiflora M. A. Curtis in Boston Jour. Nat. Hist. I. 100 (Enum. PI. 

 Wilmington) (1834), not Linnaeus. 1 Groom, Cat. PI. Newbern, 21 (1837). 

 Wood & McCarthy, Wilmington Fl. 34 (1886). 



1 I have seen in the herbarium of the Academy of Natural Science at Phila- 

 delphia a specimen collected by Curtis at Wilmington and named by him A. nudi- 

 flora, which proves to be R. atlanticum. The true R. nudiflorum has not been found 

 near Wilmington. 



