130 THE AZALEAS OF NORTH AMERICA 



Small in North American Flora gives "Savannah River'' as the type locality 

 for his A. lutea, but if A. lutea of Linnaeus is taken up as the name for R. calendu- 

 laceum it could be based only on Cadwallader Colden's enumeration of the plant 

 from southern New York and not on the locality given by Michaux for var. ». of 

 his A. calendvlacea. Unfortunately Michaux*s specimen collected on the Savannah 

 River near Two-Sisters Ferry and cited by him under a. flammea does not belong to 

 A. calendvlacea as described by him, but to R. speciosum, and therefore the speci- 

 men labeled in his herbarium /3. crocea must be considered the type specimen and 

 the high mountains of Carolina or more particularly the Blue Ridge Mountains 

 of North Carolina, the type locality. Though it is against general usage to take 

 var. 0. as the type of a species which is divided into varieties, the specific descrip- 

 tion in this case appears to be chiefly based on the specimen designated var. (9., 

 particularly the character "corolla tubo hirsuto, laciniis breviore" must have been 

 taken from var. /?., for in var. a. (R. speciosum Sweet) the tube is distinctly longer 

 than the lobes and less conspicuously hirsute, also the color "floribus . . . calen- 

 dulaceis" referring to Calendula which has yellow to orange flowers, indicates the 

 mountain form, var. p., rather than var. a., which has bright red to scarlet flowers. 



Rhododendron calendulaceum does not show much variation except in the color of 

 the flowers and in the length of the sepals, which are usually oblong, but sometimes 

 rather small; the style is normally pilose near the base, but sometimes glabrous, 

 as in Harbison's No. 171 from Macon County, North Carolina, in Small's speci- 

 men from White Rock Mountains, Smyth County, Va., and in Britton's & A. M. 

 Vail's specimen from the same locality. There are also forms with partly salmon- 

 colored flowers of which I collected a specimen myself on Roan Mountain of North 

 Carolina, and I have before me a specimen from North Carolina stated to have 

 pink flowers. The two following forms have been described, of which the first 

 represents the typical form, though probably more color forms could be distin- 

 guished but this cannot be done satisfactorily without copious living material of 

 undoubtedly spontaneous origin. 



Rhododendron calendulaceum f. croceum Sweet, Hort. Brit. ed. 2, 

 343 (1830), as var. o. — Rehder in Mitt. Deutsch. Dendr.Ges. XXIV. 

 (1915), 225 (1916), as forma; in Bailey, Stand. Cycl. Hort. V. 2942 

 (1916), as var. 



Azalea calendvlacea /3. crocea Michaux, Fl. Bor.-Am. I." 151 (1803). — Sims in 

 Bot. Mag. XLI. t. 1721 (1815). — Rehder in Bailey, Cycl. Am. Hort. I. 

 121, (1900) .» 



Flowers yellow to orange-yellow with a darker blotch, sometimes partly 

 salmon-colored. As stated before there are all kinds of intermediate shades be- 

 tween the plants with scarlet flowers and those with yellow flowers. 



Rhododendron calendulaceum f. aurantium Rehder, comb. nov. 



Azalea calendvlacea a. Ker in Bot. Reg. II. t. 145 (1816), not var. a. flammea 



Michaux. 

 Azalea aurantiaca Dietrich, Darst. Vorz. Zierpfl. 4, t. 1 (1803). — Hort. apud 



Dumont de Courset, Bot. Cult, ed 2, III. 333 (1811), as var. of A. nudiflora. 

 ? Azalea cocdnea major Dumont de Courset, I. c. (1811), as var. of A. nudiflora. 



1 Additional Reference to Coloeed Plate: Mag. Bot. & Gard. II. 119, 

 t. 36, fig. 7 (1834). 



