most of the pedicels 1-2 lines long, flowers in dense panicles, 

 some stem leaves broader and even oval to reniform and ses- 

 sile. This is a very characteristic species and varies but little, 

 growing in sand in the Tropical Life Zone. Specimens of this 

 m my herbarium are my No. 3826 from the Needles, Califor- 

 nia, May 5, 1884, a specimen from southern Utah collected by 

 F. M. Bishop many years before in Powell's expedition down 

 the Colorado, also specimens from Panamint Canyon, May 3, 

 1897, and Death Valley, California, May, 1906, all collected by 



Gilia scopulorum Jones was always confounded with G. 

 inconspicua by Gray, but it is easily separated from that by 

 the color of the leaves alone, their softness and pubescence 

 and also by the peculiar flowers. The leaves are incised 

 after the style of G. incisa. The flowers are remarkable for 

 the varying length of the tube on the same plants, which go 

 from not longer than the calyx to four times as long. This is 

 shown well by my specimens from Shepherd's Canyon, Cali- 



Gilia aggregata var. maculata n. var. 



Tall and virgate, green; flowers large, about V/z inches 

 long, obovate-lanceolate lobes nearly half the whole, beauti- 

 fully mottled with red and yellow and white even in the bud 

 and with short and filiform tips. Soldier Canyon. Sierra 

 Madre Mts., Chihuahua, Mexico, at 6000 feet alt., Sept. 16, 

 1903. 



Gilia phamaceoides var. Harknessii (Curran Bull. Cal. 

 Acad. 1 12 as species). This cannot be maintained as a spe- 

 cies. It is very common in the Great Basin. 



Like G. filicaulis, but flowers shorter than the calyx lobes 

 and stamens not exserted ; neither bracts nor calyx lobes di- 

 lated. The type of this is the specimen from Salmon Mea- 

 dows, Idaho, at 4000 feet altitude, July 22. 1899. To this I 

 also refer my No. 2465 from Soda Springs, California, July 23, 

 1881 ; also those from Emigrant Gap, California, June 30, 1882; 

 and Colfax. California, July 3, 1882. 



Gilia Schottii, setosissima and Mathewsii are only vari- 

 eties of one species, though they appear to be remarkably dis- 

 tinct, for thev have all sorts of inter-grades. 



Gilia debilis Gray. This has been one of the rarest of all 

 the genus and has been given several names. The writer de- 



