258 Rhodora o [DECEMBER 
Lawrence, were naturally very perplexing since it would be surprising, 
at least, to find B. heterodoxa upon an inland lake of Connecticut. 
Consequently, plans were made for the writer to join Mr. Woodward 
in a further field-study of the colony. Unfavorable weather, however, 
forced the abandonment of this plan and on September 21, 1915, 
Mr. Woodward alone visited the station and collected an abundant 
series of specimens, many sheets of which have been generously 
supplied to the writer; and these plants in all their characters agree 
with the earlier collections. 
A close study of the material shows it to have exactly the achene 
of B. heterodoxa and to be consistently two-awned. A check-study 
of B. connata and its varieties shows that in that species all well 
developed central achenes of the heads are consistently 4-awned and 
with the highly developed mid-ribs becoming almost wing-like in 
maturity. It would appear, then, that the Pocotopaug Lake material 
must, at least for the present, be placed with B. heterodoxa, although 
it is geographically remote from the type region of the latter species. 
In this geographic isolation, however, the plant is comparable with 
many other characteristic species of Prince Edward Island, the 
Magdalen Islands and eastern New Brunswick, which are outlying 
representatives of austral types isolated by hundreds of miles from 
the nearest known stations to the south,! and it is probable that 
further exploration, especially in the coastwise strip of southern New 
England, will reveal colonies of B. heterodoxa in the intermediate area. 
In typical B. heterodoxa the leaves are either simple or 3-5-parted, 
the blades of the simple leaves or the terminal lobes of the cleft ones 
being narrowly lanceolate to narrowly ovate and very sharply (almost 
jaggedly) serrate. In the Connecticut material, however, the leaves 
as above stated, strongly resemble those of B. connata, var. petiolata, 
or, in some cases, typical B. connata, the simple leaves being oblong- 
lanceolate to ovate-lanceolate and rather bluntly dentate, the cleft 
leaves with the terminal lobe of this form; and in all the plants the 
leaves are on very elongate slender petioles. Furthermore, the 
Connecticut material has the flowering branches highly developed 
but very short in the axils, most of these branches being much shorter 
1 For example: Carex varia Muhl., abundant on Prince Edward Island but unknown nearer 
than Hancock Co., Maine; Rumezx persicarioides L. (See St. John, Ruopora, xvii. 80) of the 
lower St. Lawrence, Quebec, Prince Edward Island, and eastern Massachusetts; and Aster 
subulatus Michx., var. obtusifolius Fernald, Ruopora, xvi. 61, of northeastern New Brunswick, 
there representing A. subulatus, which reaches its northern limit in southern New Hampshire. 
