166 THE AZALEAS OF NORTH AMERICA 



strigillose midrib, and slightly reticulate. The flower-clusters are large and many- 

 flowered; the sepals semiorbicular and minute; the corolla-tube is about 3 cm. long, 

 villose and glandular-hirsute; the setose hairs of the ovary are tipped with purple 

 glands. Part of No. 36 consists of a branch with the leaves narrower and puberu- 

 lous above and short-villose beneath. Whether the numbers 17 and 18 collected 

 in bloom on March 29 belong here is doubtful. The leaves are half -grown, glauces- 

 cent on both sides and glabrous except the strigillose midrib beneath, the corolla is 

 less pubescent and the tube somewhat more gradually dilated at the apex. It has 

 some resemblance to R. atlanticum, but the corolla is less glandular and more vil- 

 lose, the winter-buds are pubescent and the habit is apparently different. Some 

 specimens from Lee and Dallas Counties, Alabama, which I have referred to typi- 

 cal R. viscoswn, may belong rather to this variety than to R. viscosum. 



Aiton's Azalea viscosa y. fissa (Hart. Kew. I. 203 [1789]. Rhododendron vis- 

 cosum t.fissum Sweet, Hort. Brit. ed. 2, 344 [1830]) described as "floribus albis ad 

 basin usque divisis, foliis saturate viridibus lucidis " is apparently not a form with 

 a regularly five-parted corolla like his A. nudiflora rj. partita, for Dumont de Cour- 

 set describes the same form under Azalea serotina N., A. fissa H. K. (Bot. Cult. ed. 2, 

 III. 336 [1811]), as follows: "Le tube de la corolla est plus court; le limbe se fend 

 quelquefois dans une des ses divisions jusqu'au tiers ou la moitiS du tube, meme 

 jusqu'a sa base; dans plusieurs corolles il ne se fend pas." Such split corollas may 

 be seen occasionally and seem to occur in certain individuals more frequently than 

 in others; for instance Harbison's No. 656 of var. montanum shows a large per- 

 centage of more or less deeply split corollas, but this can be considered only as an 

 abnormality and not as a taxonomic form. 



A form with double flowers is mentioned by Dumont de Courset in 1814 (Bot. 

 Cult. VII. 166) as "Azalea visqueuse a fleurs doubles," and in 1864 Kirchner (in 

 Petzold & Kirchner, Arb. Muse. 479 [1864]) enumerates an Azalea viscosa flare 

 pleno. I have seen neither a plant nor a specimen of a double-flowered true R. vis- 

 cosum; the forms I have seen under this name were apparently of hybrid origin. 

 There is also an Azalea glaucaflore semipleno Jager (Ziergeh. 118 [1865]). Azalea 

 viscosaglauca sousv&T.flore pleno Mouillefert, Arb. Arbriss. II. 1021 [1897]. ? Azalea 

 glauca plena Hort. Kew. [in herb. Arnold Arb.]. Under the last named synonym a 

 specimen was collected by G. Nicholson in Kew Gardens in 1880; the leaves are 

 glaucescent beneath and the corolla has the shape of that of R. viscosum, but it is 

 almost destitute of villose pubescence, the sepals are well developed and oblong 

 and the young branchlets are almost glabrous except stipitate glands. These 

 characters, together with the early flowering time (June 1) when the leaves were 

 only partly developed, suggest R. atlanticum. 



Rhododendron arborescens Torrey, Fl. U. 8. 425 (1824). G. Don, 

 Gen. Syst. III. 847 (1834). Loudon, Arb. Brit. IV. 1145 (1838). 

 Gray, Syn. FL II. 1, 40 (1878). Sargent, Gard. & For,!. 400, fig. 64 

 (1888). Dippel, Handb. Laubholzk. I. 416, fig. 269 (1889). Coul- 

 ter & Watson, Gray's Man., ed. 320 (1890). Robinson & Fernald, 

 Gray's New Man. ed. 7, 631 (1908). Schneider, III. Handb. Laub- 

 holzk. II. 500, fig. 329 1-m, 330 b (1911). Millais, Rhodod. 115 

 (1917). 



? Azalea viscosa Marshall, Arbust. Am. 15 (1785), not Linnaeus. 

 Azalea arborescens Pursh, Fl. Am. Sept. 152 (1814). De Candolle, Prodr. 

 VII. 2, 716 (1839). Gray, Man. 268 (1848); ed. 5, 299 (1867). M. A. 



