Robinson: Polycodium 553 



emphasis on articulated pedicels as a means of distinguishing the 

 Thibaudieae from the Vaccinieae but, unfortunately, has gone 

 far beyond the facts. 



One character remains to which Professor Greene has called 

 particular attention, the open aestivation of the corolla. All 

 evidence that has been obtained confirms his statement on this 

 point for every species of Polycodium, including V. Kunthianum. 

 This does seem of such importance that the genus may properly 

 be maintained; the other characters previously mentioned may 

 be treated as collateral, by one who is dealing with the species of 

 America north of Mexico, but it must be remembered that not 

 one of them can be relied on to distinguish Polycodium from all 

 other genera, even within the limits of North America. 



Rafinesque* published Polycodium thus: "67. The species of 

 Vaccinium with campanulated corollas, must form a peculiar 

 genus or subgenus, Polycodium. In fact the whole tribe of 

 Ericacea or Bicornia must be newly modelled." And this is all, 

 with one very important exception. The paragraph occurs in a 

 review of Pursh's Flora Americae Septentrionalis and may fairly 

 be interpreted by reference to that work. Pursh divided Vac- 

 cinium^ primarily into species with deciduous leaves and those with 

 evergreen leaves, dividing each of these in turn on the basis of 

 campanulate as contrasted with urceolate corollas. His species 

 with deciduous leaves and campanulate corollas were V. stami- 

 neum, V. album, V. arboreum, V. dumosum, V. frondosum, and V. 

 pallidum; those with persistent leaves and campanulate corollas 

 were V. Vitis-Idaea, V. myrtifolium, and V. crassifolium. Vac- 

 cinium stamineum thus comes first, and V. album Pursh is regarded 

 by most authors as the same species. Polycodium, therefore, may 

 be held to be typified by Vaccinium stamineum L., but Rafinesque's 

 genus in its entirety was a mixture of widely differing elements, 

 including representatives not only of Batodendron but of Gay- 

 lussacia and of different sections still included in Vaccinium by 

 nearly all authors. 



Picrococcus of Nuttall is based almost entirely on Vaccinium 

 stamineum L. (including V. elevatum Banks & Soland.) with the 



*Am. Monthly Mag. 2: 266. 1818. 

 t Fl. Am. Sept. 1: 284-290. 1814. 



