336 CARNEGIE INSTITUTION OF WASHINGTON. 



flower under normal and under changed conditions. The competition 

 of Rosa and Ruhus for polhnators was studied in much detail, the results 

 indicating that Ruhus is about twice as efficient in its attraction. In 

 a number of flowers, the life histories were again worked out in detail 

 and recorded graphically. 



Field and Garden Study of Genera and Species, hy F. E. Clements and H. M. Hall. 



The general study of taxonomic criteria has been continued in both 

 field and garden. The field study has been chiefly confined to the 

 genera Chrysothamnus, Artemisia, Atriplex, and Haplopappus, though 

 Pentstemon, Castilleia, SoUdago, Bouteloua, Aristida, and Quercus have 

 received some consideration. A particular endeavor has been made 

 to discover the species of the first four in their type localities, and to 

 determine the various forms as well as the range of variation. Statis- 

 tical studies have been made of a number of species and it is expected 

 that all will ultimately be included. Special field studies were made at 

 Pike's Peak and in the vicinity of Colorado Springs and Caiion City, 

 and much of the time of the field expedition from Colorado to Oregon 

 and California was devoted to this work. The garden studies of the 

 Madiece were extended by plantings at Tucson, while control experi- 

 ments were begun on such paired species as Petalostemon candidus 

 and purpureus, Psoralea tenuifiora and argophylla, to determine both 

 their plasticity as to new forms and their reversibility as to each other. 

 Much attention has been given to the details of a taxonomic system 

 which both recognizes and reveals evolutionary relationship, and to a 

 system of nomenclature which will be both usable and attractive. 



Climax Formations, hy F. E. Clements. 



The constant study of the climax formations of the West has made 

 possible the first complete though necessarily summary treatment of 

 them in "Plant Indicators." As in the past, continued attention has 

 been paid in the field to the climax formations and associations, and 

 their successional development. Major consideration has again been 

 given to the formations of widest extent, namely, grassland, sagebrush, 

 desert scrub, and chaparral. The last two were studied more or less 

 intensively at Tucson from October to May, and the grassland and 

 sagebrush chiefly in Colorado and westward to Oregon in the summer 

 and autumn. A detailed investigation was made of the grouping and 

 alternation of the dominants of the Larrea-Franseria desert scrub 

 within a radius of 75 miles of Tucson during the winter, and of the 

 dominants of the Larrea-Prosopis association in southern New Mexico 

 from May to July. The ecotone between the desert-scrub and grass- 

 land was traced in central New Mexico, as well as the transition from 

 the desert plains to the short-grass plains and mixed prairie. This 

 latter association has come to be recognized as the most important of 



