STANDLEY — ALLIONIAOEAE OF MEXICO. 397 



KEY TO THE SPECIES. 



Perianth about 55 mm. long \ %0m gmndiflora. 



Perianth 35 mm. long or less. 



Leaves strongly undulate, very unequal at the base; lobes of the 



perianth deeply cleft 2.0. hypogaea. 



Leaves entire, equal at the base; lobes of the perianth shallowly 



retuse - 3.0. rosei. 



1. Okenia grandiflora Standley, sp. nov. Plate 76. 



Stems stout, viseid-villous, the internodes 5 to 8 cm. long; leaf blades oblong or 

 oblong-ovate, 17 to 52 mm. long, 17 to 45 mm. wide, thin, rounded or obtuse at the 

 apex, unequal and truncate or subcordate at the base, with strongly undulate margins, 

 glabrous or sparingly pilose above, villous beneath along the veins, the opposite 

 leaves strongly unequal; petioles stout, 20 to (JO mm. long, densely villous and 

 viscid; involucre of 3 triangular-ovate bracts 3 mm. long, very viscid and villous; 

 perianth 55 mm. long, the limb 40 mm. wide, the tube very narrow at the base but 

 soon widening above, the limb rather deeply lobed, sparingly villous without; 

 stamens about 15, unequal, strongly curved upward; peduncles 55 to 60 mm. long at 

 anthesis, soon lengthening to 20 cm. or more, stout, strongly villous; fruit not seen. 



Type in the Gray Herbarium, collected by Mr. C. G. Pringle in the barranca of 

 Tequila, in the State of Jalisco, July 4, 1893 (no. 5444). Apparently the plant was 

 not obtained in quantity, for this collection has not been seen in any other her- 

 barium. In addition to being strikingly different in the size of its flowers and lesa 

 obtrusively in some minor particulars, such as the number of stamens, this plant 

 occurs in a very different habitat from that of 0. hypogaea, the latter being invariably 

 maritime. 



The photograph is of the type specimen. 



2. Okenia hypogaea Schlecht. & Cham. Linnaea 5: 92. 1830. Plate 75, A. 



Type localit y, "In collibus arenosis prope Vera Cruz." Type collected by Schiede 

 and Deppe (no. 107). 

 Specitnens examined: 



Oaxaca: Chiltepec, 1890, Rovirosa 726. 



Vera Cruz: Type collection in the Bernhardt ilerbarium of the Missouri Botan- 

 ical Garden; 1910, Orcutt 3293. 

 Colima: Acapulco, 1S95, Palmer 490; Colima, 1897, Palmer 39 and 40. 

 Sinaloa: Altata, September 2, 1904, Brandegee; sand dunes along the beach 

 north of Mazatlan, April 4, 1910, Pose, Standleij & Russell I400(j. 

 Although first collected so long ago, no member of the Allioniaceae has been so little 

 collected or is so poorly represented in collections. In some of the larger herbaria 

 no material at all is to be found. In general appearance specimens simulate Wede- 

 liella, and possibly they have been neglected by collectors who confused them with 

 these common plants. 



The writer has examined a specimen of the type collection in the herbarium of 

 the Missouri Botanical Garden. It accords well with specimens collected along the 

 coast in various parts of Mexico. Unfortunately, no fruit is present. Because of 

 the peculiar fruiting habit of the genus, exactly like that of the peanut, the fruit is 

 usually left underground when the plants are pulled up. The authors of the genus 

 seem to have had fruit, for they describe it as having ten longitudinal ribs. If this 

 statement is true, the plant of the Pacific beaches is different from that found on the 

 eastern coast, for its fruit is smooth. Doctor Heimerl states that he has seen no fruit 

 of the genus in any European herbarium. 



In April, 1910, the writer found one or two growing plants on the dunes of pure 

 Band that line the coast north of Mazatlan, Sinaloa. These were merely individuals 

 that had persisted from the previous season, the plant being an annual. The dead 



