Harper: Explorations ix Georgia 1904 4^9 



stations, quite different from the type-locality as well as from 

 each other. On April 16, I found a few specimens beginning to 

 flower {no. 21 jo), around a spring in the rocky woods mentioned 

 above, about two miles south of Elko in Houston County. On 

 May 3. I collected it again in low woods near Thalmann, Glynn 

 County, where it was more abundant (no, 2iSj). Here it was 



associated with 



# 



ru^/ 



bnrmtni nuduin^ and some other species which seemed strangely 



• incongruous ; while at the Houston County station some of its 



n ea res t n e i gh b o rs we re Hepatica^ Sa ngidnaria^ Gera niu fu , and 



other familiar northern plants. The three known localities have 



probably at least one feature in common however, the absence of 



the Lafayette formation and consequently more or less calcareous 

 soil. 



The specimens of these later collections have culms ascending 

 or nearly erect (not nodding as in the type), and shorter and less 

 densely tufted than in the original specimens ; no tendency to 

 proliferation in the inflorescence was observed (though this may 

 develop later in the season) ; and the type specimens were evi- 

 dently somewhat abnormal in these respects. The new specimens 

 seem to be just as distinct from related species as the old ones were. 



Cladium mariscoides (Muhl.) Torr. 



This is probably common enough in the glaciated region of 

 the North, but very k^^ stations are known for it in the South.* 

 Consequently I was rather surprised to find a good deal o{ it in a 

 grassy cypress pond between Pinehurst and Unadilla in Dooly 

 County on May 21 {jw. 2221). It was then not quite in flower. 



The leaves of this species are different from any others known 

 to me, and I have never seen them correctly described. Toward 

 the base they are channeled, and a cross seciion there would show 

 a shallow trough with flat bottom and erect sides. Tapering up- 



^vard, the leaf gradually folds inward on its midrib, and within an 

 ^nch or so of the apex the margins (remaining about the same 

 '^vidth) become completely united, giving there a triangular cross 

 section. 



* 



See Rhodora i : 43, 98, 204; 2: 123, 202; 5: r33 ; ^- '^8 ; 7: 72; also 



Mohr, Contr. U. S. Nat. Herb. 6 : 410. 1901 



