152 ENVIRONMENTAL CONDITIONS. 



The present treatment is a purely geographical and climatological 

 study, wherein it is sought to determine the approximate distribution 

 of different cHmatic types in the United States. Since the aim of 

 Part III of this publication will be to study the comparative distribu- 

 tion of vegetation types and climatic areas, it has seemed desirable 

 to prepare many of our climatic charts upon a base-map showing by 

 different patterns the distribution of the great general vegetational 

 areas of the country. The discussion of the climatic ranges of the 

 several vegetational types will be reserved for Part III. 



The climatological charts themselves have been prepared, as far 

 as possible, directly from the particular set of data involved, in the 

 same manner as is the common practice in climatological work in 

 general. The numerical data were first placed upon a copy of the 

 Relief Map of the United States, of the U. S. Geological Survey, 

 (17 by 27f inches, hypsometrically colored and also furnished with 

 contours for altitude), the numbers being written near the positions 

 of the respective stations. Then the isoclimatic lines were sketched 

 in in pencil and the map laid away for a time. At a later date it 

 was worked over a second time and changes made that seemed to bring 

 the lines nearer toward expressing the probable truth. 



In this revision the topography of the country was constantly 

 scrutinized and the contour-lines of the base-map were allowed to 

 influence the course of the isoclimatic lines in many instances, especially 

 where the stations for which data were at hand were too far apart to 

 show the true directions of these lines. The preparation of charts 

 of this sort is at best largely a matter of guesswork; information is 

 not available for the plotting of climatic details. For meteorological 

 purposes it is usually quite undesirable to have charts showing such 

 details, but for our purposes they were quite essential. It thus be- 

 came evident early in the present studies that the method commonly 

 used in the drawing of meteorological charts — of carrying isoclimatic 

 lines from plain to plain directly over high mountain ranges, with 

 little or no attention to altitude— should not be resorted to here; 

 at least it should not be used on so grand a scale as is conomon in 

 meteorological work. Furthermore, the degree of approximation to 

 the actual truth is surely often greatly increased by a due regard to 

 topography, rather than by an almost blind following of inadequate 

 climatic data modified only by a desire for lines as smooth as possible. 

 If we are aiming at as true a picture of natural conditions as is at- 

 tainable, it is obviously more undesirable to pass a given line over 

 a locality where we are absolutely certain it should not pass than to 

 draw it through some other area where it surely does pass and wherein 

 only its proper placing is questionable. Thus, if there be given two 

 stations with the same climatic datum, on either side of a range of 

 mountains, it may be taken as certain that the conventional joining 



