484 PAPERS ON GENTIANE. 
In the Gila Valley, Rothrock, 1874 (325), and southeastward into Mexico, Gregg, ete. — Stems 1-14 feet high, 
the tallest of our species; leaves 1-1} inches long, distinguished by its large bi- or tri-colored flowers with acutish 
lobes. E. venusta, Gray, with which it has been confounded, is a ents plant with larger deeper-colored flowers, 
broader obtuse corolla-lobes, and usually longer anthers and larger seeds. —The anthers of the different species of 
Erythrea are of different shape and size, from orbicular and oval to oblong and linear and 3-4 or 5 mm. in length; all 
become at last spirally twisted after they have shed their pollen, the longer more conspicuously so, the shorter nich 
less. The stigmas of this genus have often been misunderstood, probably because mostly observed in dried and 
pressed specimens. They are never capitate or funnel-shaped, but always bilobed. Before maturity, they remain 
closed, and only after the anthers have shed their pollen do both halves separate and spread out, just as the Gentians 
behave. In the form of the stigma I find valuable characters for the grouping of the species, and especially for the 
distinction of the American ones from those of the Old World. The stigmas of the former are flabelliform and 
broader than long; those of the latter are orbicular-ovate or oblong to linear ; shortest in EH, spicata and linearifolia, 
and longest in major, where they are twice as long as wide, and in maritima, in which the length is 3 or 4 
times as great as the width. £. australis from hewealia, as well as Cicendia from Europe, have stigmas like [190] 
the Avieriean species, but B. Chilensis (at least a specimen collected by Ochsenius in pica which I refer to 
that species) differs from all the others in having elongated, cuneate, emarginate stigm 
a are identical with those of a large group of enim. hick comprise perhaps all of 
Grisebach’s Chironiew and Chlorew, being very numerous and very small, from 4 to scarcely more than 3 mm, in the 
longer oo and, when fully mature, globose or oval, with their surface favose-reticulate. 
orth American genera belonging to this group may be arranged thus : — 
1. Stamens inserted in the tube : ERYTHR#A. EvsToMA. 
2. Stamens inserted in the throat: SaBBaTiA. MuicRocaLta. — Catora, Carronta, and Ses of the Old 
World also belong here. 
My investigations have convinced me that the genus eons, or at least its typical and original species, C. pusilla, 
is nothing but a reduced form of Erythrea, distinguished from it only by its small oval anthers, less than 0.5 mm. in 
length, too short to twist much, but still, when drying, sometimes twisting a little. Its stigma is by no means ca 
tate, but regularly bi-lamellate, and, much like that of the American species, ig T NO URE Hit and broader oc 
long. The flowers are 4-parted, which often occurs in genuine specie of Erythrea ; seeds 0.4 mm. long, large for the 
size of the plant. It will have to bear the name Erythrea pusilla. 
RYTHRHZA Dovatasil, Gray, Fl. Calif. 1, 480; Syn. 113.—Slender, a span to a foot high, loosely and pani- 
culately branched, usually sparse-flowered ; leaves from oblong to linear, acutish ; flowers on strict, slender pedicels, 
4-5 lines wide ; lobes of the pale pink corolla obtuse, much shorter than the diate: anthers usually only 1 mm. long, 
style short, stigma about 1 mm. wide ; seeds subglobose, 0.4 mm. in diameter. — Arizona, Utah, and northwestward 
to Oregon. This plant has been confounded by Mr. Watson with his #2. Nuttallii, which, however, is a smaller and 
more leafy plant, with larger flowers and much larger seeds 65 mm. long), but much smaller stigmas. 
GEnTIANA! HUMILIS, Stev.; Griseb. in DC. Prod. 9 ; Engelm. Trans. Acad. St. Louis, 2, pl. 9, figs. [191] 
1-5; Gray, Syn, 120. — Biennial, with large, broadly oak: St white-margined basal leaves, and few or 
1 The presence or absence of folds or plaits between the of the = of St. Louis, vol. 2, pl. 7, 8, 9, and 11, ver- 
lobes of the corolla and the mode of attachment of the anthers satile anthers are erroneously represented as turning indis- 
to the filament separate the Gentians into two large — tied intone outward or inward. This is a mistake, as the 
natural sections, already recognized by old authors: Gen above account of the living action of the anthers shows. In 
tianella, Borkhausen, Gray, has a corolla without folds and the figures of @. humilis and prostrata, pl. 9, the anthers 
the ‘nthers versatile ; Pnreumonanthe, Necker, Gray, are also figured as versatile in that unnatural manner, while 
corolla with folds between the lobes and fixed anthers, a in these species they are constantly erect and introrse, as 
will not be useless to explain the term versatile in respect to well before opening as when effete. 
anthers, as many seem to misunderstand it, so that they In Praeumonanthe the anthers remain fixed in two forms. 
speak of versatile anthers as accidental and unconnected with In one section, comprising mostly smaller plants, with 
a physiological process. The fact is that in Gentianel/a the smaller flowers (@. prostrata, verna, Altwica, humilis, utricu- 
ops introrse in the bud and after it first opens; but losa, etc.), they are introrse. In another section, the true 
as the flower is fully expanded (generally toward the perennial large-flowered Pnewmonanthes, to which we must 
middle of the day) the anthers gradually assume a horizontal add also an annual, G. Douglasiana, and the European 
position (the eee base raised and turned toward the as cruciata, ~— are extrorse. In the genera Halenia, Plew- 
yet immature and closed stigma), open the cells upward, and royyne, Swertia, and Frasera, all represented by plants col- 
begin to shed their pollen. Toward evening, the now effete lected in ewe Gaioldiins, we find the same avaapement 
is turned over backward, and on the next morning we of versatile anthers as in Gentianella. It therefore seems 
find it hanging on the back of the sacri the notched roper to enumerate, first, the Gentians with fixed a anthers, 
lower end turned up and the empty cells directed outward. and next those with versatile anthers, and then, joining 
Thus in about twelve hours it ian Gesestting an almost com- them, the other genera with similar versatile antheral 
plete circle. In my figures of Gentians in the Transactions ment. 
