ECOLOGY. 315 



extensive and diverse, but also because it is most subject to modification by- 

 animals, the latter are most abundant and most readily visible in it, and the 

 correlations and reactions between the plants and animals are most evident. 

 The desert also offers some of the same advantages, though the number of 

 animals is much smaller. This was strikingly exemplmed in Death Valley, 

 where few were to be found in the chmax areas, while the hydrosere in the 

 several ponds at Saratoga Springs yielded more than 20 species of birds and a 

 corresponding number of other animals. For somewhat similar reasons the 

 insects, birds, and rodents are the animal groups that have received first 

 consideration. At the same time an endeavor has been made to project into 

 the past the results and principles wrought out in the biome of to-day, in the 

 hope of refining the methods and conclusions of paleo-ecology. A compre- 

 hensive survey of the hterature on the animal side has been initiated with 

 the object of bringing together all the results of definite ecological nature 

 which bear on the problems of the biome. 



Climax Formations, hy F. E. Clements and E. S. Clements. 



Practically all the associations of the 9 climax formations which occupy 

 the West have been traversed in the course of the 12,000 miles traveled by 

 motor during the year, the majority of them receiving both intensive and 

 extensive study for periods of several weeks to a month or more. This has 

 permitted a closer and more direct comparison of the associations of each 

 formation and a more detailed scrutiny of the ecotones between the various 

 chmaxes than has been possible heretofore. One result of fundamental im- 

 portance has been to broaden the relationship between the associations and 

 to enhance the unity of the chmax. This has been most striking in the case 

 of the grassland formation, because of its greater extent and diversity, but 

 it applies with almost equal force to the scrub and forest chmaxes. 



The grassland associations have been studied almost through their entire 

 extent. The bunch-grass association was examined from the Mexican border 

 through California and Oregon to northern Washington and thence eastward 

 to Utah. This revealed greater unity than was known to exist between the 

 southern and northern portions and made it possible to trace the contact 

 between the bunch-grasses and the tall-grasses which formerly dominated 

 the northeastern portion of the Great Basin. While previous observations 

 indicated that the bunch-grass prairie once extended to western Wyoming, it 

 was found that this was true only of the foothills, and that the plain proper 

 was covered with such tall-grasses as Stipa comata, Agropyrum glaucum, and 

 Sporobolus cryptandrus. The behavior of the short-grasses and sedges was 

 studied along their western frontier and a mass of additional evidence was 

 obtained as to the universal conversion of the mixed-prairie climax into the 

 short-grass subchmax in response to overgrazing. The mixed, true, and sub- 

 chmax prairies and the intervening ecotones were examined from Colorado 

 through central Nebraska to Kansas and Missouri, and the subchmax prairie 

 was then traced to southeastern Texas to its contact with the coastal prairie 

 on the south and the southern mixed prairie on the west. In western Texas 

 the latter passes rather abruptly into the desert plains, which extend well 

 beyond central Arizona, finally yielding to the true climax of desert scrub. 



The desert-scrub formation was again examined through the Colorado and 

 Mohave Deserts and in addition its structure and development were studied 

 throughout Death Valley and the adjacent valleys to the east. This cor- 



