MINT FAMILY 411 



Var. revoluta Munz. Compact shrub similar to the species; leaves somewhat dimorphic, 

 the leaf -blades of the primary leaves narrowly oblong, 2 to 3 inches long, 2% to 3 lines mde, 

 revolute-margined, shortly petioled or subsessile, those of the leaves fascicled in the leaf axils 

 revolute-linear, % to 2 inches long, 1 line wide, both sorts remarkably bullate, shining green and 

 glabrous above, tomentulose beneath ; flowers essentially as in the species ; corolla lavender, the 

 hairy transverse band in throat broader than in the species. — Steep canon sides, 5 to 800 feet: 

 Santa Eosa Isl. May. 



Tax. note. — Salvia mellifera var. revoluta is narrowly endemic on Santa Eosa Island. A com- 

 pact bush, it is, by virtue of this compactness, different in appearance from the species (Euth 

 Hartwell). The variety is, however, but slightly kno'\\Ti ; knowledge of it is in the formative stage. 

 The notes here set down rest mainly on flowering material from a bush cultivated in the Santa 

 Barbara Botanic Garden (Van Eensselaer 1029) and on the tj'pe collection (which is without 

 flowers). The flowers resemble those of Salvia mellifera rather closely: the calyces are alike in 

 the two; the corollas are much alike save that the tube in Salvia mellifera var. revoluta is some- 

 times slightly longer ; the ring of hairs in the corolla-throat is broader in var. revoluta than in 

 Salvia mellifera (where it is very narrow), but this feature is variable. The vestige of the lower 

 arm of the stamen connective is exactly like that in the species, only somewhat more reduceil. "The 

 stamen length appears to be more or less uniform on any one plant but varies from one plant to 

 another. On one indi%'idual the anthers are well included ; on another individual they are exserted 

 nearly half the length of the upper lip." — Field note on Santa Eosa Island by Eeid Moran, 1941. 



With its dimorphic leaves, var. revoluta differs from the species most strikingly in two features 

 relating to the leaves: 1. The intervals of the reticulate venation in both sorts of leaves in var. 

 revoluta are so strongly bullate or inflated as to resemble markedly the neatly-fitted pieces of 

 a tesselated pavement. The venation in the species, on the contrary, is impressed only in such 

 degree that the upper leaf surface may be described as rugose but not as tesselate, save occasionally 

 on the margins of the leaves. 2. In var. revoluta the leaves of the axillary fascicles as a whole 

 differ strikingly in shape from the primary leaves and differ markedly from the characteristic leaf 

 of the species. The most definitely established differences between the mainland and insular shrubs 

 thus rest on the leaves — characters which, in view of the close similarity of the flowers, may not 

 seem wholly sufficient to justify specific status. The range of variation in the flower of Salvia 

 mellifera covers essentially the features found in the flower of var. revoluta. 



Eef 3.— Salvia mellifera Greene, Pitt. 2:236 (1892) ; Jepson, Fl. W. Mid. Cal. 459 (1901), 

 ed. 2, 359 (1911), Man. 870, fig. 816 (1925). Audihertia stachyoides Benth., Lab. Gen. et Sp. 313 

 (1833), type from Cal., Douglas; not S. stachyoides Kunth (1817). Audibertiella stachyoides 

 Briq., Bull. Herb. Boiss. 2:73 (1894). Eamona stachyoides Briq., I.e. 440 (1894). Var. jonesii 

 Munz, Bull. S. Cal. Acad. 26:24 (1927), type loc. Ensenada, L. Cal., Jones. S. mumii Epling, 

 Madrono 3:169 (1935), type loc. Hamilton ranch, L. Cal., Epling 4" Sobison. Var. revoluta 

 Munz, Bull. S. Cal. Acad. 26:23 (1927). Audihertia stachyoides var. revoluta Bdg., Proc. Cal. 

 Acad. ser. 2, 1:216 (1888), type loc. Santa Eosa Isl., T . Brandegee : not S. revoluta E. & P. (1798). 

 S. brandegei Munz, I.e. 21:69 (1932); Epling, Ann. Mo. Bot. Gard. 25:117, pi. 20 (1938). 



Salvia bernardina Parish; Greene, Bull. Cal. Acad. 1:211 (1885), type loc. San Bernardino, 

 Parish. S, columbariae var. bernardina Jepson, Man. 869 (1925). Shrub 1 to 2 feet high; leaves 

 shallowly pinnatifid, the divisions not extending more than halfway to the midrib ; bracts small ; 

 upper lip of calyx tipped with 3 spine-like teeth, or these united to the end into 2 or 1. — Canons 

 on the north side of the San Bernardino Valley, 1000 to 2500 feet: San Bernardino, Parish 4383; 

 Lytle Creek and Cajon Canon (Bull. S. Cal. Acad. 26:27). In spite of previous varietal reference 

 of this form to Salvia columbariae, examination and pondering of authentic material leads now 

 to the consideration of relationship to S. mellifera. Probably there is no close connection with 

 S. columbariae. Doubtless the explanation of this enigmatic plant is to be found in hybrid origin 

 as suggested by P. A. Munz (Bull. S. Cal. Acad. 26:26). 



11. S. leucophylla Greene. Chaparral Sage. San Luis Purple Sage. Stems 

 woody below, leafy, white-farinose above, 3 to 4 feet high ; leaf-blades obloiig-lance- 

 olate or the lowest ovate, obtuse, the upper with truncate base, finely rugulose above, 

 white-tomentulose beneath, 1 to 3 inches long, short-petioled ; flowering branches 

 with 3 to 5 very dense head-like flower-whorls ; bracts oval or oblong, densely white- 

 farinose; calyx splitting down in front, at length emarginate behind, a-s if composed 

 of two lateral lobes; corolla pink or light purple, 5 to 7 lines long, the tube hardly 

 exserted ; stamens and style much exserted; connectives almost continuous with the 

 fertile filaments. 



Dry often barren chaparral slopes, 100 to 2500 feet : Monterey Co. to Orange 

 Co. May-June. 



Field note. — In the central part of its range, as in Santa Barbara County, tliis shrub, Salvia 

 leucophylla, forms colonies of three, five or ten acres extent in the hills, growing on steep westerly 

 or northwesterly slopes. Where the conditions for it attain Iheir optimum, it develops pure stands 

 sharply delimited against grassland. It also occurs as individuals scattered through the chaparral 



