318 CARNEGIE INSTITUTION OF WASHINGTON. 



in the Southwest, where under a minimum rainfall a small decrease produces 

 a large effect. Likewise, all communities have been modified to some extent 

 by grazing, but grassland and especially those associations in an arid climate 

 have been changed the most, while the reverse is true of cultivation in the 

 West. This is universal only in humid or subhumid regions, and its maximum 

 effect is consequently to be found in the subclimax and true prairies. Fire, 

 on the contrary, has but relatively little effect on grassland, while it modifies 

 forest profoundly and often scrub also. In the communities changed by each 

 of these processes the relicts have their own peculiar significance, though all 

 are in harmony with the general principle. This is to the effect that the 

 relict plant or community is one that has enjoyed a certain degree of protec- 

 tion against the destruction due to changes of climate, to grazing, cultivation 

 or fire, and naturally also to other minor destructive processes. The persis- 

 tence of such relicts is due to the fact that plants once thoroughly established 

 have a much better chance of adjusting themselves to changing conditions 

 than seedlings do, the hazards for practically all species being far the greatest 

 during the first year. Hence the study of the changes in vegetation that are 

 occurring or have occurred in the last few thousand years is first of all a search 

 for protected places or regions in which former dominants or subdominants 

 have been able to persist. In the case of climatic changes these are valleys, 

 sandhills, and higher altitudes during the dry phase and exposed slopes and 

 ridges during the wet. They are primarily areas fenced against cattle in the 

 case of grazing, and those untouched by plowing, pasturing, or building in the 

 case of cultivated regions. Fire relicts regularly result from changes of wind 

 or weather, differences of density, slope, land-form, barriers, nature of the 

 fire, etc. These afford explicit evidence upon which the reconstruction of past 

 vegetation can be based, current changes followed, and future ones depicted. 



Rainfall and Climatic Cycles, by F. E. Clements. 



The investigation of the rainfall of the western United States and Canada 

 in relation to the sunspot cycle has been continued during the year. The 

 tables for more than 1,000 stations are being checked against the new edition 

 of the section summaries of the Weather Bureau, and it is hoped to publish 

 the results of the study soon after all these have become available. In 

 order to afford a larger number of r-ecords extending back beyond 1860, some 

 of the longer records of the East are also being taken into account. Rainfall 

 records from widely separate countries in the other five continents are like- 

 wise being assembled in order to ascertain whether they exhibit cycles and 

 whether these bear any relation to the sunspot cycle. 



In the endeavor to evaluate some of the factors in rainfall and to discover 

 a possible connection with rainfall cycles, a preliminary series of measurements 

 has been made of the water-loss from natural vegetation and cultivated crops 

 in midsummer. These have been carried out at Lincoln, Nebraska, in the 

 true prairie with a rainfall of 29 inches; at Phillipsburg, Kansas, in the mixed 

 prairie with a rainfall of 23 inches; and at Burlington, Colorado, in the short- 

 grass plains with a rainfall of 17 inches. While climatologists have more than 

 once shown that the planting of crops in the prairie States has not resulted 

 in an increased rainfall (cf. Loveland and Swezey, The rainfall of Nebraska, 

 Bull. 45, Nebr. Agr. Exp. Sta., 1896), the opinion still prevails that rainfall 

 has increased with cultivation and is continuing to do so. The results 

 show that one native species, Andrcpogon furcatus, transpired 27.6 pounds 

 per square foot of sod during a 15-day period, while a similar area of oats 



