272 LIVINGSTON— CLIMATIC AREAS [April i8, 



When these indices are placed upon a map and isocHmatic lines 

 are drawn in the usual way, we have the chart which is shown in 

 full lines in plate IX. ^^ The data are in terms of hundredths of an 

 inch per day and their range of magnitude is from less than two 

 to over sixteen hundredths. This is not a proper place for detailed 

 discussion, but it is at once obvious that the precipitation lines of 

 this chart tend strongly to take a north-south direction, thus crossing 

 our isothermal lines and dividing the country into irregular climatic 

 areas each of which might be defined by the use of these two systems 

 of lines. As has been stated, the data from which both temperature 

 and precipitation charts have been constructed are relatively very 

 satisfactory, and it may be surmised that the combination chart here 

 presented is fairly reliable as a general picture of the climatic condi- 

 tions of the country as measured according to the method here set 

 forth. 



3. Amount of evaporation during the frostless season. — Russell's 

 data on evaporation in the United States are for but a single year, 

 and that not a calendar one. The probability of error introduced 

 by assuming these data to be normal is very great, yet, as has been 

 emphasized, these are the only data yet available, and we must either 

 employ them or follow the custom of our Weather Bureau and 

 ignore the important subject of direct evaporation measurements 

 entirely. More to illustrate the value of evaporation records than 

 with any thought that the details of the present study may be free 

 from large error, I present here the results of an approximate deter- 

 mination of the mean depth of daily evaporation for the frostless 

 season. It is to be noted that the data for the earlier months of the 

 frostless season are from the summer of 1888, while those for the 

 later months are from that of 1887, an unsatisfactory state of affairs 

 made necessary by the exigencies of Russell's study. 



Russell's published data are given by months, and, since the 



" It is to be remarked that this and the two following charts attempt no 

 more than an approximation to normal conditions. The lines are so placed, 

 however, as to represent the data as these have been obtained. Where no 

 stations are present topography has been used as an indication of the prob- 

 able position of the lines. All of the data here employed will be published 

 elsewhere. 



