554 ROBINSON: POLYCODIUM 
addition of a second species, Picrococcus floridanus, certainly 
congeneric with the former, of which more hereafter. 
Professor Greene is unduly critical of Nuttall for overlooking 
or disregarding such a publication as Rafinesque’s of Polycodium 
for the genus containing Vaccinium stamineum, and Nuttall was 
not the first offender. In 1836 Rafinesque himself has the fol- 
lowing:* ‘‘ADNARIA Raf. fl. lud. probably a subgenus of the 
Codorolla or Vacciniums with bell flowers, which see.” But 
Codorolla does not appear again, and Rafinesque did not consider 
this reference worth indexing. 
It is apparent that the adoption of Polycodium as a generic 
name, to be typified by Vaccinium stamineum L.., is barely justified. 
Before leaving the subject of generic subdivision in the Vac- 
cinium alliance it is desired to amplify a statement above made 
with regard to the importance of characters derived from the 
stamens. The anthers may be 2-awned or awnless. Ina solitary 
specimen, Wright 2202, referred to Vaccinium Ramonii Griseb., 
some anthers were found with a single central awn. Further 
examination showed that this was not a constant character, even 
within a single flower; but none of the anthers were awnless. 
The value of anther awns as a diagnostic character obviously 
depends upon the degree of their constancy, and final judgment 
will be influenced, consciously or unconsciously, by the nature of 
the grouping thus achieved. 
Examination of the flowers of every species but one, 
within the limits of North America, as well as of some extra- 
limital material, gives this result. There are 3, or more likely 
only 2, species, V. meridionale Sw., of Jamaica, V. consanguimeum 
Klotzsch, of Panama and Costa Rica, and V. multiflorum Benth., 
of Colombia, in which this character has to be handled with 
extreme caution. All of these are very closely allied; indeed, 1t § 
doubtful if the first two can be held distinct. In all of these the 
awns are very delicate and often closely appressed, to such an 
extent that they might escape detection, unless considerable care 
be taken in their search. Further, in V. meridionale they are 
often so reduced that they are practically wanting, yet otner 
H rt 
stamens in the same flower may possess them, slender and shor 
Re AONE SO ES 
found 
* New FI. 1:65. 1836. 
