9G Rhodora [May 



former of these two extends to Newfoundland, 1 but west of Nova 

 Scotia reaches its northeastern limit in the lower Penobscot valley; 

 while 8. atlanticum has heretofore been unknown northeast of southern 

 York County, Maine. On open gravelly soil Pease and Linder had 

 also found a plant which so closely matches 8. arenicola of the sands 

 of New Jersey, Long Island and Nantucket that there can be little 

 question as to its identity. The Yarmouth material, however, seems 

 like a starved 8. gramineum with the short and stiff basal fibres (one 

 of the chief characters) persistent perhaps through a response to 

 ecological conditions, while material which Pease, Long and I sub- 

 sequently found on dry plains at Middleton, Annapolis County, 

 seems like 8. angusfifolium except for the stiff and persistent tufts 

 of basal fibres. May it not be that 8. arenicola, instead of being a 

 true species, is an ecological state due to the sandy substratum in 

 which it grows? 



But still more interesting was the discovery that the spruce bogs, 

 besides having the plants one would naturally expect (the boreal 

 Carex pauperculaf C. pauciflora, Smilaeina trifolia, Yaccinium, Oxy- 

 coccus, Empctrum nigrum, etc.), shelter along with the already well 

 known coastal plain Carrx ailaniica Bailey (C. stcrilis of the Manual) 3 

 and C. cxilis, the delicate little southern C. Ilowei Mackenzie, 4 the 

 plant treated in the 7th edition of Gray's Manual as C. scirpoidcs, 

 var. capillacea but clearly a distinct species of the coastal plain. 

 C. Hotcei, which extends in New England north to the lower Merri- 

 mac, is from Cape Cod southward one of the dominant plants of the 

 so-called Louisianian and Carolinian Cypress (Chamaecyparis) 

 swamps, but throughout western Nova Scotia it is quite as dominant 

 a sedge of the "Hudsonian" spruce swamps (fig. 1). Another 



'See Bicknell, Bull. Torr. Bot. CI. xxvii. 238 (1900) and Fernald. Am. Journ. 

 Bot. v. 243 (1918). 



2 There seems no good reason to recognize vars. irrigua (Wahlenb.) Fernald and 

 pallcns Fernald. Fifteen years of fleld-work since they were proposed shows them 

 to be only trivial variants. 



' C. stcrilis Willd. has been variously misunderstood, but Mackenzie (in Britton & 

 Brown, 111. Fl., ed. 2, i. 377) seems to have reached a satisfactory solution of its 

 identity: a very distinct but little-collected species of limestone regions from New- 

 foundland and Anticosti westward to Minnesota, and south through the limestone 

 region of western New England to northern New Jersey, Pennsylvania, etc. This 

 plant, until recently merged with C. interior Bailey (C. scirpoidcs, at least of my own 

 treatments), differs from it in having very rough beaks which barely exceed the 

 broad and very long brown scales. The coastal plain plant which I have called 

 C. sterilis is mostly C. atlanlica Bailey. 



* Mackenzie, Bull. Torr. Bot, 01. xxxvii. 245 (1910). 



