108 
of this study I have myself collected and of each made descrip- 
tions of fresh corollas, and noted other features to be gained only 
in the field. The importance of such work in taxonomic study 
needs emphasis. 
In the present revision keys are given to the genera and species. 
These are detailed for points of definite contrast. These keys 
apply only to the species of our flora, and the warning must be 
made that the generic and tribal contrasts may be of little or no 
assistance beyond this territory. But just such keys as these 
are of most value to the local worker, and moreover it is by 
combining such analyses from various regions that we may hope 
ultimately to build more thorough family keys. An inductive 
process! 
For each genus the type-species is stated. For each native 
species information of its type is stated, quoted from the original 
describer. This includes the statement of the particular speci- 
men from which the first description was made and of the place 
of its collection. The later history of each name is traced. 
Extra-limital synonyms, even if the names have been current 
here, are not included except by brief mention. But all names 
ever proposed based upon plants occurring native in this area are 
supposed to be included. : 
With respect to distribution I should like to undertake a study 
for which the data at hand in our herbaria is not yet sufficient. 
Moreover my own observations have not as yet been sufficiently 
prolonged over this area. The counties best represented in 
herbaria are those of Connecticut; .New York, from the High- 
lands southeastward, including all Long Island; New Jersey, 
with considerable gaps to the northwest; Pennsylvania south- 
east of the Blue Ridge; and northern Delaware. Northwest 
of the Highlands and of the Blue Ridge botanical collections 
have been few and scattered, the regions best known being the 
Pocono Plateau of Pennsylvania, and sections of Ulster, Greene 
and Delaware counties, New York. 
Dr. Witmer Stone, in his Plants of southern New Jersey, has 
traced with a master-hand the distribution of vegetation for the 
Coastal Plain portion of that state. That regions of as sharp 
