STUDIES IN PORTULACACE^. 89 



Lewisia rediviva, var.? Yosemitana. — Caudex verv 

 short; flesh}- root ver}' slender: leaves succulent, linear 

 or spatulate, 1-2 cm. long: pedicels cylindrical, ^-i cm. 

 long, jointed just below the flower and crowned by three 

 ovate scarious bracts : sepals 2, broad, concave, emarginate 

 or more deeply notched at top : petals 5, about 2 cm. long, 

 exceeding the sepals: stamens about 15: style rather 

 shortly cleft; ovar}- circumscissile from a broad base; 

 seeds too young for description. The flowers fall from the 

 jointed pedicel, in drying, even more pi-omptly than in 

 the type. — Collected somewhere about Yosemite Valley, 

 by Mrs. Willie F. Dodd, in the summer of 1891. 



This plant has so much the aspect of a depauperate 

 L. rediviva that in lack of mature fruit, I prefer to de- 

 scribe it as a variety of that species. Whether it prove 

 to be distinct or not, it is plainly related more closely 

 to L. rediviva than is any other species, and effect- 

 ually breaks down the remaining barrier between Lewisia 

 and the Lewisioid section of Calandrinia. There re- 

 mains then only to consider whether all these species shall 

 be united with Lewisia on the common characters of the 

 circumscissile capsule and persistent style, or whether 

 Lewisia shall be merged into Calandrinia. I think the 

 first would be the more convenient. Dr. Gray says that 

 circumscissile dehiscence occurs in some South Am- 

 erican species of unlike habit, but does not specify which. 

 There are in the herbarium of the CaHfornia Academy 

 about thirty species from that region, none of which 

 seem to be circumscissile, though it must be confessed 

 that several of them are too 3'oung to admit of certainty. 



If Montia and Claytonia are both to be maintained, 

 it must be on the lines laid down by Mr. Howell — all 

 3-ovuled species remanded to the first, and those with 6 



