146 Rhodora [July 



appropriately quoted: "Much less so [unstable] are two other char- 

 acters which, indeed, seem to be almost always sharply distinctive 

 although they have been little emphasized in descriptions. In S. 

 tteliarit the main leaves, broadest at or above the middle, are dis- 

 tinctly narrowed to the base and the usually acute apex, and the entire 

 plant, unless carefully pressed, readily turns black in drying. Sab- 

 batia campanulata, on the contrary, shows little or no discoloration 

 on the herbarium sheet, and the commonly obtuse leaves, linear, 

 linear-oblong, oval or, low on the stem, actually ovate, are broadly 

 sessile or suhclasping." 



Besides the excellent distinctive characters pointed out by Bicknell 

 there are some others which seem equally important. S. stellaris 

 is a plant of saline or brackish marshes and all the specimens examined 

 by the writer from throughout the range, from Massachusetts to east- 

 ern Mexico, are annual (or possibly biennial). S. campanulata, on the 

 other hand, is at least in Massachusetts a plant of fresh sandy pond- 

 shores or sphagnous bogs, and such labels from the southern states 

 as indicate habitats read: "low, grassy pine barrens and meadows," 

 "dam]) pine barrens," etc., indicating that the New England habitat 

 of the plant is not unique. Furthermore, all the material of S. cam- 

 panulata which has been carefully collected shows it to be a slightly 

 caespitose perennial with a short subligneous rhizome. The majority 

 of specimens, merely picked or pulled, have no roots but seem to have 

 been broken from a crown. One other character, not so constant, but 

 fairly reliable, is in the bracting of the lateral peduncles. In well 

 developed S. gteUaris all the peduncles are naked; in well developed 

 S. campanulata the lateral peduncles usually bear 1 or 2 median bracts. 



As already emphasized Sabatia campanulata is a plant of the fresh 

 sandy or peaty shores and marshes. In New England it is very local, 

 seen by the writer only from three regions of Massachusetts: Nan- 

 tucket Island; the borders of Mary Dunn's Ponds in Barnstable on 

 Cape Cod; and a single station, Pembroke in Plymouth County, where 

 it was collected on September 10, 1SS4, by W. L. Foster. It was 

 reported by the late Alfred "YY. Hosmer x from a single station in Con- 

 cord, Massachusetts, but the writer has not seen Mr. Hosmcr's 

 material. Its presence as an indigenous plant in Concord is open to 

 some question owing to the transplanting activities. of the late Minot 

 Pratt, but the occurrence in Concord or adjacent towns of such 



i Rhodora, i. 224 (1899). 



