` 190 Rhodora [OCTOBER 
the foliage, and the type of roughness on the angles of the culm are 
often helpful characters. The length of the anther is useful in some 
cases, though highly variable within rather wide limits and differing 
greatly according to condition, that is to say whether fresh and full 
or dry. This length for the various species is as follows: C. albursina 
1.4-2.0 (2.2) mm. dry or fresh; C. blanda 2.0-2.8 mm. dry, 3.0-3.5 
mm. fresh; C. laxiflora 2.0-3.0 mm. dry, 3.0-3.5 mm. fresh; C. ormo- 
stachya 2.8-3.2 mm. dry; C. crebriflora 2.0-2.2 mm. dry; C. ignota 
2.9-3.2 mm. dry; C. anceps 2.0-3.0 mm. dry, 3.2-4.0 mm. fresh; C. 
striatula 3.2-4.3 mm. dry; C. styloflexa 3.0-4.0 mm. dry; and C. 
leptonervia 1.3-2.3 mm. dry, 1.7-2.5 mm. fresh. The anthers at the 
very summit of the spike are usually much smaller than those farther 
down, and the measurements given do not apply to these. The 
length and plumpness of the perigynia may vary greatly, frequently 
without the variation being of taxonomie value. Often the filling 
out of the achene seems to broaden and shorten the perigynium. 
The term beak as here used signifies a point with concave sides in 
distinction from a merely acute apex. Many specific characters 
fluctuate greatly, and though generally true, occasionally fail, so that 
the species are best defined by the sum of all the characters. On the 
whole the nine species here admitted, though very closely related, 
are distinct. In two or three of the southern species there is an appar- 
ent tendency to produce stolons, but the specimens are few and 
imperfect. Much more field study is necessary to understand prop- 
erly the southern forms. The writer wishes to take this opportunity 
to protest against the wretched labels so common in herbaria. In 
the case of several of the southern species treated in this paper, it 
has been impossible to determine whether they are plants of wet or 
dry soil, clay or sand, shady or exposed places, as not a single label 
bears so much as a suggestion of such facts. The localities, too, are 
for the most part imperfectly given. 
The oldest specific name connected with the group is that of Lam- 
arck, C. laxiflora. The application of the Lamarckian name has 
always been a matter of doubt. By the earlier authors the name was 
applied to narrow-leaved forms of what is here called C. anceps Muhl., 
but usually included also C. leptonervia and C. striatula Michx. Boott's 
C. laxiflora was more especially C. striatula. Bailey in his later 
papers, after inspecting the specimens in Lamarck's herbarium, 
1Mem. Torr. Bot. Club i. 32 (1889). 
