oblanceolate and tapering into a slender petiole; all of them hirsute with white spreading hairs: flowers axillary from the base up; calyx tube very short, the lobes lance-oblong, 3-4mm long; petals broadly ovate-oblong, tridentate at the nearly truncate summit, about twice as long as the calyx-lobes, exceeding the longer stamens and about equalled by the pistil: capsules purplish, small and slender, less than 2cm long, variously flexed and somewhat angled: seeds small, pale, smooth, usually oblique at base and obliquely pointed at apex.

I name as type A. D. E. Elmer's no. 3192, Monterey, Cal., distributed as S. micranthum. No. 5099, by C. A. Purpus, seems to be the same.

10. S. HIRTELLUM (Greene) Small, Bull. Torr. Bot. Club 23:190; Oenothera hirtella Greene, Fl. Francis. 215.-LÉVEILLÉ in his monograph reduces this to a form of S. micranthum, but this does not seem to be justified by his specimens.

11. S. MICRANTHUM (Hornem.) Walp. Repert. 2:77; Oenothera micrantha Hornem. Hort. Hafn.-That LÉVEILLÉ should take up the much later name O. hirta, and then reduce to this species such distinct forms as S. hirtellum, and S. Nelsonii, seems a little strange. However, he has described a very good variety which may be written

11a. S. MICRANTHUM Jonesi (Lévl.), n. comb.-Taking as the type the first number cited, viz. Hansen's 543, Amador Co., Cal. 1892, I would name as a close duplicate C. C. Parry's specimens (in the Missouri Botanical Garden) simply labelled "Oenothera. June 1889. Cal." Blanche Trask's Avalon specimen, cited by LÉVEILLÉ, seems rather to belong with the species itself. It is quite probable that the species as it now stands is an aggregate.

11b. S. MICRANTHUM exfoliatum, n. var.-Branched from the base, the stems stoutish, the bark white, shreddy, and exfoliating in thin sheets, giving the plant the appearance of S. decorticans: pubescence of the stems ciliate, that of the leaves and fruits closer and somewhat appressed: capsules sharply angled, contorted.

I cite here the following specimens: C. R. Orcutt, Colorado Desert, April, 1889; C. A. Purpus, no. 5083, Erskine Creek, Cal., 1897.

12. S. BISTORTUM (Nutt.) Walp., Repert. 2:77. Oenothera bistorta Nutt., T. & G. Fl. N. A. 1:508.

12a. S. BISTORTUM Veitchianum (Hook.), n. comb.; Oenothera bistorta Veitchiana Hook., Bot. Mag. pl. 5078.-The characters

which were supposed to distinguish this variety sufficiently from S. bistorta and to constitute it a species become less well-defined the larger the series of specimens. Even the greater length of capsule and beak seems to be a variable quantity.

13. S. SPIRALE (Lehm.) Fish. & Mey., Ind. Sem. Hort. Petrop. 2:50; Oenothera spirale Lehm. in Hook. Fl. Bor. Am. 1:213.— Assuming it to be a fact that this Californian plant is distinct from S. cheiranthifolium of South America, the specific name as given by LEHMAN in Hooker's Flora is the next available one.

13a. S. SPIRALE viridescens (Lehm.), n. comb.; Oenothera viridescens Lehm. in Hook. Fl. Bor. Am. 1:214.-If this be a good species, it certainly is very difficult to distinguish from the preceding. It was given only varietal rank by WATSON in his revision (Proc. Am. Acad. 8:592), under the name suffrutescens-the woody base and possibly perennial duration, with the somewhat larger flowers, being the characters that he used. But even in these respects the species and the variety seem to grade into one another.

13b. S. SPIRALE clypeatum (Lévl.), n. comb.; Oenothera clypeata Léveillé, Monog. Oenothera, 222.-Distinguishable by the broad, shield-shaped leaves which are densely canescent, and by the large flowers (often 4cm across).

14. S. DECORTICANS (H. & A.) Small, Bull. Torr. Bot. Club 23:191; Gaura decorticans H. & A. Bot. Beech. Voy. 343.-SMALL seems to be well within bounds when he assigns priority to the name of HOOKER and ARNOT. The species, though apparently greatly variable, is so merely before it begins to blossom when quite small, at which time it is smooth and erect. With age it becomes large, more spreading, and roughened with the loosened shining shreddy bark. WATSON'S characterization of the seeds as "cellularpubescent" is a good one.

15. S. tortum (Lévl.), n. sp.; Oenothera chamaenerioides torta Lévl., Monog. Oenothera, 230; O. alyssoides minutiflora Wats., Proc. Am. Acad. 8:591.-Branched from the base and spreading, becoming at length nearly prostrate; leaves glabrous, mostly basal, oblanceolate and tapering into slender petioles: capsules about 2cm long, cinereous, variously contorted.

LÉVEILLÉ is right in allying this with S. chamaenerioides, but on habit alone, to say nothing of the fruits, it is entitled to specific rank.

Following are specimens illustrating: Jones, 5548, Manti, Utah; Nelson, 4691, Granger, Wyo.; Cusick, 2515, Malheur River, Oregon; Trelease, 4435, Shoshone, Idaho; Nelson, 4707, Green River, Wyo.; Godding, Milford, Utah.

15a. S. TORTUM Eastwoodae, n. var.-Leafy throughout, the leaves oblong-linear: flowers very small: capsule tapering into a slender beak, spirally coiled at base.

This is probably a good species. Only one specimen is before me: Alice Eastwood, Grand Junction, Colo., May, 1892.

16. S. TORTUOSUM A. Nels., Proc. Biolog. Soc. 17:95. 1994; Oenothera gauraeflora caput-medusae Lévl., Monog. Oenothera, 226. -LÉVEILLÉ'S plate shows only a single branch from Lemmon's specimens.

17. S. UTAHENSE Small, Bull. Torr. Bot. Club 23:191.-Whitened with a tomentose pubescence, branching from the base upward, 15cm or more high (the plants in hand are all young): leaves ovate, obovate, or oblanceolate, 2-4cm long, generally tapering to a short petiole: flowers crowded in terminal somewhat corymbose racemes, white; calyx-tube longer than the lanceolate lobes; petals obovate, 5mm long, longer than the stamens but surpassed by the pistil: capsule linear, less than 2cm long, more or less contorted.

The specimens before me were collected by L. N. Goodding at Milford, Utah. As the original description calls for yellow flowers I thought at first Goodding's specimens were another species, but agreeing in most other respects I am going to assume that "flowers yellow" was a clerical error.

18. S. BOOTHII (Dougl.) Small, Bull. Torr. Bot. Club 23:191; Oenothera Boothii Dougl., Lehm. in Hook. Fl. Bot. Am. 1:213.Seemingly seldom collected. Typical specimens by L. F. Henderson, Shoshone Falls, Idaho, July, 1897. Many of the specimens referred

to this species belong to the next.

19. S. Lemmoni, n. sp.--Branched from the base up, 2-3dm high, stem and branches rather stout, crinkled-hirsute; branches divaricate-ascending: leaves rather large, variable in size, 2-5cm long, oblong or broader, mostly acute at apex, lower tapering into petioles, hirsute-ciliate especially beneath: flowers in a crowded terminal short-hirsute raceme, lengthening into a bracteate fruiting spike; calyx tube but slightly enlarged upward, scarcely as long as its narrowly lanceolate lobes; petals broadly obvate or suborbicular,

about 7mm long, slightly longer than the calyx lobes and stamens, equalling the style; stamens similar and equal: capsule slender, tapering to summit, ascending, somewhat bent or contorted, about 2 cm long.

This has passed as S. Boothii Dougl. Similar as the descriptions seem, the two plants are quite distinct in appearance. S. Boothii is glabrous and with shreddy bark on the older stem; it branches mainly near the base, the branches also branching; its flowers are much smaller, and the capsules are shorter and more contorted.

The type is J. G. Lemmon, no. 103, eastern flank of Sierra Nevada, Cal. 1875. Two good specimens, both in herb. Mo. Bot. Garden.

20. S. Hitchcockii (Lévl.), n. sp.; Oenothera gauraeflora Hitchcockii Lévl., Monog. Oenothera, 226.-Softly hirsute or ciliate, branched from the base, 15-25cm high; branches slender and more or less paniculately branched above: root leaves oblong, irregularly dentate, about 3cm long, tapering into a petiole one-half as long; stem leaves smaller, bract-like, sessile, broadly linear or lanceolate: flowers crowded in bracteate secund racemes; calyx tube slender, scarcely enlarged at summit, as long as the linear-lanceolate lobes; petals white, obovate, 3-4mm long, scarcely longer than the calyx lobes and the stamens; style slender, longer than the petals: capsules slender, 12-18mm long.

This very excellent species rests upon two specimens in herb. Mo. Bot. Garden. One bears the data "Simpson Park, July 6th, 1859 (?)," and in pencil "nothing like it known to me;" the other is blank, but both look as if they were from the same collection.

21. S. ALYSSOIDES (H. & A.) Walp., Repert. 2:78; Oenothera alyssoides H. & A., Bot. Beech. Voy. 340.

21a. S. ALYSSOIDES MACROPHYLLUM Small, Bull. Torr. Bot. Club 23:192; Oenothera alyssoides villosa Wats., Proc. Am. Acad. 8:591.

The following are, so far as the writer knows, still unknown except from the original specimens and descriptions. So far as one may judge from descriptions, they are valid and will no doubt again come to light.

S. ANGELORUM (Wats.), Oenothera angelorum Wats., Proc. Am. Acad. 24:49.

S. PTEROSPERMUM (Wats.); Oenothera pterosperma Wats., King's Rep. 112.

S. RUTILUM (Davids.); Oenothera rutila Davids., Erythea 2:61.

COOPER'S COLORADO COLLECTIONS.

In the summer of 1904, Mr. WILLIAM S. COOPER, a student in Alma College, Michigan, spent some weeks in Colorado collecting in the vicinity of Estes Park and upon Long's Peak. He secured over 300 numbers, many of them of great interest. The following I will characterize as new:

Oreocarya pulvinata, n. sp.-Cespitose-pulvinate, practically stemless, the small cushions a few centimeters across and about 1cm high; flowers as well as the leaves involved in the soft villous pubescence: leaves crowded, broadly linear, less than 1cm long: flowers few, glomerate at the summit of the reduced stems (the stems scarcely rise above the matted leaves): calyx-lobes linear, nearly equaling the corolla tube: corolla white; its tube dilated, subspherical, about 2mm long, the broad throat only partly closed by the conspicuous crests; the lobes of the limb suborbicular, about as long as the tube: stamens small, included, inserted near the middle of the tube; filament almost wanting: style short, rather thick, equaling the stamens.

This species so closely simulates Eritrichium aretioides (before the flower stalks of that species have developed) that one would almost certainly pronounce it an Eritrichium at the first glance. The pubescence and pulvinate habit are similar, but a glance at the flowers does not leave one in doubt very long.

The type material, no. 278, is very scanty, but so characteristic a species cannot be ignored. Collected on Mummy Mts., Estes Park, Aug. 12, 1904, alt. 12-13,000.

Chrysopsis Cooperi, n. sp.-Whitened with soft loose long-villous pubescence throughout: stems low, spreading, more or less decumbent at base, 10-15cm high, leafy throughout: leaves narrowly oblanceolate, tapering into a margined petiole-like base, from 2-5cm long, middle and upper stem leaves usually longer than the basal: heads solitary, terminal and axillary; terminal head large, 12-14mm high and considerably broader, subtended by some foliar bracts which are long-ciliate on the margins; axillary heads reduced downward, on successively shorter leafy peduncles, usually only the 2 or 3 uppermost developing, the others becoming sessile and aborted