157 



Delphinoa Ross, Boll. R. Orto Bot. Palerra. 1 : 117. 1897. 

 Type, I), gracillima Rosa, 1. c. 



This is also Agave potosina Robinson &, Grcenman, Proc. Am. Acad. 29:393. 1894. 

 The above genus was described as having the lower dowers in pairs whilo the 

 upper ones are solitary. 



SYNOPSIS OF THE NORTH AMERICAN SPECIES OF NISSOLIA. 



The genus Nissolia was established by Jacquin in 1760 (Enum. PI. 



Carib. 7). 



In 1825 A. I)e Caudolle (Prodr. 2 : 257 to 250) monographed the genus, 

 describing seventeen species. Of these, however, only three belonged 

 to the genus Nissolia proper, viz, W./ruticosa, N. hirsuta, and If. race- 

 mosa, the latter two being there described for the first time. 



In 1859 G. Bentham (in Mart. PI. Bras. 15, pt. 1, pp. 76, 77) reduced 

 these three to Jacquin's original N.fruticosa, and in addition described 

 the species N. platycarpa. 



Previously to Bentham, Dr. Gray (in 1852) and Dr. Torrey (in 1850) 

 had each described a species under Ohaetocalyx. 



In 18o'l (Journ. Linn. Soc. 5:25, 20) Dr. Gray transferred these two 

 species to Nissolia, recognizing four in all, viz, N. fruticosa, N. plaiy- 

 carpa, N. unslizeni, and N. schottii. 



In the Biologia Centrali- Americana Mr. llemsley lists five species, 

 one being without specific name, and the" specific -name hirsuta being 

 twice used on different authority. On the same page A 7 , schottii and 

 N. icislizeni are retained in Ohaetocalyx. 



Later, Watson described two additional species from Mexico, viz, 

 N.platy calyx and A r . confertifiora. All the above species except N. race- 

 mosa have been reported from Mexico. A 7 , hirsuta, which was suppressed 

 by both Bentham and Gray, appears to be distinct from ISf.fruticosa, 

 to which they* referred it. The type came from Guanajuato, Mexico, 

 and I have recently had specimens from the type locality. A T . conferti- 

 flora, it seems, should be referred to this species. N. setosa lirandegee 1 

 and A 7 , confertifiora laxior Robinson are other names which have recently 

 beeu published. The species are difficult to recoguize, but the following 

 key seems to separate them fairly well : 



EXPLANATION OF FIGURES. 



In all the following figures (Nos. 17 to 27) the same letter applies to the same part. 

 Thus, fig. a is always the legume; fig. b, the flower; fig. c, the banner; fig. il, the 

 wing; fig. e, the keel; fig. f, the stamens; fig. g, the ovary. The fruit is natural 

 size; the flowers and flower parts are enlarged twice. Fig. 17, X, msUseni, is made 

 from a specimen collected by C. G. Priugle near Chihuahua in 1885 (No. <>18). Fig. 18, 

 X. schottii, is made from a specimen collected by C. G. Pringle near Altar, Sonora, in 

 August, 1894. Pig. 19, X. platycah/r, is made from a duplicate type. Fig. 20, X, 

 print/hi, is made from the type specimen. Fig. 21, X. dirersifolia, is made from the 

 type specimen. Fig. 22, X. hirsuta, is made from specimens collected by Mr. C. G. 

 Pringle near Cuernavaca in 1886 (No. 0395). Fig. 23, X. dodgei, is made from speci- 



1 Proc. Cal. Acad. scr. 2, 3 : 127. 1891. 



