METHODS 



A survey of literature, particularly of pertinent keys and monographs, was con- 

 ducted before hybridization experiments began. In addition, some observations of floral 

 and vegetative characteristics were made during 1967 and later growing seasons. Collec- 

 tions were made of shrubs in various stages of development for illustrative purposes. 

 Artificial hybridization was tried between Cowania mexioana var. stanshuviana , Purshia 

 tvidentata, Fallugia paradoxa, and the Cowania- Purshia hybrid by treating emasculated 

 flowers of each with pollen from the others (fig. 1). Flowers which were to receive 

 pollen or to serve as controls were covered with white paper bags before anthesis 

 (fig. 1). These artificial crosses were made to determine if there are any major diffi- 

 culties in interspecific hybridization within and between these closely related genera. 



Figure 1 . — Cliffrose 

 (Cowania mexicana var. 

 stansburiana) with 

 flowers emasculated, 

 tagged, and pollinated 

 with antelope hitterhrush 

 (Purshia tridentata) and 

 Apache plume (Fallugia 

 paradoxa) pollen. 



2 



