250 Rhodora [OCTOBER 
Hemicarex, Benth.), this character, is found to become illusory by 
degrees: the margins of the glume are exceedingly thin and brought 
close together; whether they are actually connate for more or less 
than half the length of the glume appears a matter of very slight 
importance to establish a genus upon, and from the exceeding fragil- 
ity of the scarious margins it is exceedingly difficult to determine ; 
different female flowers from the same plant, treated with every care 
under water, give different results.” 
In 1887 Pax essentially followed ! Clarke's treatment, but separated 
Elyna from the Kobresia of Clarke. But in 1894 Clarke united? 
Kobresia, Elyna and Hemicarex, a course which seems far more sat- 
isfactory than the earlier one of separating them generically on illu- 
sory characters. In this treatment Clarke recognized 20 species of 
Kobresia: 13 confined to the Himalaya of northern India, occurring 
mostly at altitudes of 10,000 to 16,000 feet from Kashmir to Bhutan ; 
2 crossing the Himalaya from India to western Tibet; 1 in the 
Himalaya of northern India and western Tibet, and the Hindukush 
Range of Afghanistan; 1 extending from Tibet to Transbaikalia 
(Dahuria of Pallas); 1 from the Himalaya of northern India and 
western Tibet to Siberia and the Caucasus; and 2 of general arc- 
tic distribution, extending south in the north temperate regions to the 
Altai, Caucasus, Alps, and Pyrenees, and in the Rocky Mountains to 
Colorado. 
Of the 20 known species of Xobresia, 19 have 3-cleft styles and 
trigonous achenes, and usually (if not always) male flowers with 3 
stamens. In a single Tibetan species, A. macrantha, Boeckeler, the 
style is 2-cleft and the achene flat, not trigonous ; and for this species 
differing from all others in these two characters Mr. Clarke has pro- 
posed the sectional name Pseudokobresia. 
It is of great interest, therefore, to find in studying Carex elachy- 
carpa of the Aroostook Valley that while it has the general floral 
structure of most Kobresias it has only 2 stamens and 2 style-branches 
instead of 3, and a compressed subterete, instead of trigonous, 
achene. Thus the Aroostook Valley plant most closely approaches 
in its characters the unique Kodresia macrantha of central Asia, but 
from that species it is very clearly distinct in its elongate narrow 
! Engl. & Prantl, Nat. Pflanzenf. ii. Ab. 2, 121-122. 
* Hook. f. Fl. Brit. Ind. vi. 694-699. 
