I'lduf'ic .\ii(h'('}r!<C(i('. 177 



Eustoma andrewsii s]i. nov. 



Perenuiul rR)iii short vertii-al seiniHeshv roots with a somewhat enlarsed 

 crown or oautlex ; tlie old strins occasional I \' persisting but apparently 

 nsnally separatintr from the crown by an articulation ; stems simple below, 

 more or less fasciculately branciied above, 2-4 dm. high ; leaves from 

 elliptic-oblong below to lanceolate and acute above, mostly 8-nerved, 2-4 

 cm. long; the next year's crown leaves appearing in the autumn as rosettes 

 which are persistent and evergreen ; i)eduncles ebracteate, 3-8 cm. long ; 

 calyx deeply cleft, less than half as long as the corolla, the slender acumi- 

 nations of its lobes being two-thirds of its length ; corolla a deep purple, 

 S-4 cm. long, its tube nearly one-third of its length, its lobes elliptic-obo- 

 vate; stamens short, the filanients rather thick, anthers sagittate, erect; 

 style stoutish, scarcely longer than the ovary and shorter than the mature 

 capsule. 



My attention was recently called to this beautiful species by Mr. Andrews, 

 who pointed out some of the essential distinctions between this and E. 

 RnsKellianum (L.) Griseb. Its perennial character he tested in the field. 

 " Of several hundred plants not one had failed to produce the rosette of 

 leaves or buds for the next year's growth." Attention may also be called 

 to the smaller deep-purple corolla and the absence of peduncular bracts. 



Secured near Boulder, Colo.. 1904. I have the same from I\Ir. C. S. 

 Crandall, " Meadow at LaPorte, altitude 5,500 ft., Aug. 21, 1805." 



Pieurogyne fontana sp. nov. 



Glabrous throughout ; stems slender, simjile or with a few narrower erect 

 branches, 1-4 dm. high ; leaves linear, mostly narrowly so, thin with dis- 

 tinct midrib and two faint lateral nerves, 20-25 mm. long, the lowest soon 

 deciduous and never rosulate ; flowers in a narrow somewhat panicled 

 raceme, having long, very slender pedicels, pentamerous ; bracts foliar ; 

 sepals green, linear, resembling the bracts but shorter, usually o-nerved 

 as are also the bracts ; corolla often surpassed by the sepals, its lobes 

 elliptic-oblong, sub-acute, about 5-nerved ; stamens half as long as the 

 corolla-lobes ; the anthers oblong; mature capsule translucent, numerously 

 ovuled, ultimately as long as the sepals. 



It has been customary to call the Pieurogyne of the Rocky Mountains 

 P. rotuta. This, I think, is not justified. That species seems to skirt the 

 northern boundary of the continent, from Labrador and Greenland to 

 Alaska. I can find no mention of it in the Rocky Mountains. Rydberg 

 makes no mention of it in the Flora of Montana nor Howell in his Flora of 

 the Northwest. Macoun gives it the distribution in the British Provinces 

 previously indicated by Gray. There seems to be no good reason for the 

 statement " and south in the Rocky Mountains to Colorado." The Colorado 

 species which has passed as P. rotnta and which here characterized under 

 the name P.forilann seems to be closely circumscribed, being probably 

 confined to north central Colorado and the adjacent border of Wyoming. 



