212 



area. Along their lines of contact in southern New England, 

 our area and in northern Indiana and llHnois, this intergrades 

 somewhat with its obvious parent, A. piirpurea. 

 3. Agalinis purpurea (L.) Pennell. 



Gerardia purpurea L. Sp. PI. 610. 1753. "Habitat in 

 Virginia, Canada." The Linnean diagnosis includes 

 both long and short-pediceled plants, so could include all 

 pink ( = "purple") flowered species. The first citation 

 accompanied by a figure, Plukenet's "Digitalis virginiana 

 rubra, foliis &" facie Antirrhini vulgaris,'' evidently the 

 prevalent plant of the Atlantic seaboard now under 

 consideration, is counted as the type. 



Gerardia purpurea grandiflora Benth. in Comp. Bot. Mag. i: 

 208. 1836. "Hab. New Jersey." Type, labeled "New 

 Jersey, Torrey 1834," seen in Kew Herbarium. 



Agalinis palustris Raf. New Fl. Amer. 2: 62. 1837. "Near 

 marshes .... From New England to Carolina." Type 

 not known to exist. Evidently intended for the prevalent 

 plant of the Atlantic seaboard. 



Agalinis longifolia Raf. I.e. 62. 1837. "Near streams New 

 Jersey to Virginia." Type not known to exist. A smaller 

 form. 



Gerardia purpurea f. albifiara Britton in Bull. Torr. Bot. Club 

 17:125. 1890. New Jersey. An albino state. Plants 

 with pure white corollas are occasional in any species of 

 this genus. 



Gerardia purpurea parvula Pennell in Proc. Acad. Nat. Sci. 

 Phila. 62: 572. 191 1. "Serpentine, Wawa, Delaware 

 county, Penna., F. W. Pennell 268g, coll. Sept. 25, 

 1910, in Herb. Acad. Nat. Sci. of Phila." The'smaller- 

 flowered depauperate plant characteristic of the Serpen- 

 tine Barrens. 



Agalinis purpurea (L.) Pennell in Bull. Torr. Bot. Club 40: 

 126. 1913. 



Aureolaria purpurea (L.) Farwcll in Rep. Mich. Acad. 



Sci. 20: 189. 1918. 

 Flowering from late August to mid-September, fruiting Sep- 

 tember to October. 



