488 CORRELATION OF DISTRIBUTIONAL FEATURES. 



the entire frostless season throughout the range is meaningless. In 

 the Desert area the entire hfe-history of this annual plant is passed 

 in the earliest weeks of the frostless season and it escapes the more 

 arid conditions of the late spring and early summer by existing only as 

 dormant seeds during that season. Cephalanthus is in foliage through- 

 out its range during the entire frostless season, but, like several other 

 plants that we have used, it is found only in palustrine situations, and 

 its ability to withstand high intensities of evaporation is entirely differ- 

 ent from the ability of Covillea or of Artemisia tridentata to withstand 

 the same intensities. In short, the conditions expressed by the moisture 

 ratio are not appUcable to Cephalanthus nor to any other aquatic or 

 palustrine plant, since conditions of precipitation in the distributional 

 areas of such forms have no immediate relation to their water-supply. 

 The fact that a given plant is able to withstand a particular climatic 

 condition of high or low intensity does not mean that the plant is able 

 to withstand it in all parts of the distributional range of that intensity. 

 The eastern deciduous trees are unable to follow westward the tempera- 

 tures which are favorable to them, and Covillea tridentata is unable 

 to follow northward the moisture conditions which favor it. This is 

 simply another way of stating the fact, already emphasized, that the 

 influence of a given climatic intensity is determined by the accom- 

 panying intensities of other cUmatic conditions. And, furthermore, 

 the manner in which a given cHmatic intensity is modified by other 

 features of the cUmate is different for almost every two plants that 

 may be compared. 



In the tables of climatic extremes may be noted very many cases in 

 which plants or vegetations range through only a small part of the 

 total gamut of a particular chmatic condition. It may be seen more 

 readily in the diagrams (figs. 21 to 74) that there are numerous 

 extremely short blocks, indicating that the plant in question meets only 

 a small part of the nation-wide scale of intensities of this condition. 

 It stands to reason that a narrow amplitude of a climatic condition, 

 represented in the graphs by a short block, indicates that the condition 

 in question is more critical for this plant than the conditions which 

 show a wide amplitude. If, for example, a plant is limited on the 

 south by a boundary which corresponds closely with the isotherm of 

 60° normal daily mean temperature and is bounded on the north by the 

 corresponding isotherm of 50°, the dimensions for this element of the 

 climate will be narrow. Such a distribution, however, would cross 

 nearly all of the isoclimatic lines indicating differences in the normal 

 daily mean precipitation, so that the dimensions for that element 

 would be very broad. It is obvious that such a distribution would be 

 controlled by the daily mean temperature of 50° and 60°, and would 

 have no control from the daily mean precipitation. The narrow and 

 broad dimensions, or amplitudes, indicated respectively by short and 



