ALTAMAHA GRIT REGION OF GEORGIA 39 



the colors are somewhat generalized, purple including both the 

 common pink-purple of Rhexia, Sabbatia, and Gerardia, and the 

 deep purple of Vernonia and related genera. The dark purple,, 

 such as occurs in the heads of many Composite, is kept distinct r 

 for that probably has a different entomological significance. In 

 the case of anemophilous flowers the color is replaced by a 

 dash. And where a color or time of flowering or other 

 character is not known, or not easily described, it is simply 

 omitted. 



If space had permitted it would have been interesting to in- 

 dicate for each species in these lists its total range, its mode of 

 dissemination, and special adaptations to environment, if known. 

 But these data, when known, have been borne in mind and sum- 

 marized for each group as far as possible, the same as if they had 

 been printed. It would perhaps have facilitated reference to 

 repeat each list in systematic sequence, so as to contrast the 

 relative abundance, range, and other characters of related species, 

 and to show at a glance the number of representatives of each 

 genus and family, but this would have greatly increased the size 

 of the work. 



The names of the species are intended to be the same as in the 

 taxonomic list in the latter part of the work, where author-cita- 

 tions and other bibliographic details will be found. The phae- 

 nological data are also intended to correspond throughout- How 

 these were obtained will be explained at the beginning of the 

 taxonomic list. 



The treatment of each habitat-group ends with a partial sum- 

 mation of some of the facts brought out in the lists, together with 

 some of the characters above mentioned which cannot very well 

 be included in the lists for lack of space. The summation of the 

 times of flowering is accomplished graphically in each case, except 

 where the species are too few in number, by means of a phas- 

 nological diagram. It would require too much space to explain 

 here the method by which these diagrams are constructed, but 

 they are probably as accurate as the data on which they are 

 based, if not more so. For as the flowering periods are just as 

 likely to be overestimated as underestimated, errors of this kind 



