FOREST TYPES IN CENTRAL ROCKY MOUNTAINS. 73 



An effort on the part of the writer to overcome the difficulties due 

 to the small capacity and fragility of the container, by substituting a 

 larger metal container for the glass tubes, was also unsuccessful, 

 although this instrument was used for several months in 1915 at the 

 control station. The experiment showed at least that the volumetric 

 method with evaporimeters could never be satisfactory for year long 

 observations, because of the very great changes in volume at the 

 freezing point, and for that matter with every change of temperature. 

 Often these volume changes were sufficient entirely to obscure the 

 water loss for a period of several days in cold weather. 



As a result of this experience, four types of wick evaporimeters have 

 been devised and tested, in which evaporation losses were determined 

 by weighing. Type 4 has proved so satisfactory under all conditions 

 that an effort has been made to determine the evaporation rates for 

 all stations by means of this instrument'. Types 2 and 4 have been 

 described by the present writer (J) as have also the f undamental prin- 

 ciples on which they are based. Type 2 was put in service in April, 

 1916, and gave fairly satisfactory records at several stations during 

 that year. It was an objectionable feature of this type that rain and 

 snow were not entirely excluded from the water chamber, although 

 in the so-called " shade" instrument of each pair this was partly 

 accomplished. The records of this instrument are, therefore, pre- 

 ferred to those of the "sun" instrument, which might absorb the 

 precipitation of an area about one-twentieth as great as the exposed 

 wick. The several instruments in use in 19 10 were never standard- 

 ized, but this objection was in part overcome by rotation among the 

 several stations. 



The usable evaporation record, then, begins witn 1916, and a large 

 part of it has been secured during the years 1919, 1920, and 1921. 

 At one station (F-13), however, where types 2 and 4 were not used 

 for a whole year, it seems advisable to complete the record for 12 

 months by reference to the data obtained by the use of type 1 instru- 

 ments. This type will not be described, except to say that its evapora- 

 tion surface was a cotton wick, fully exposed to the air and sun, in a 

 horizontal plane. Like the Piche and Livingston instruments, it 

 gave relatively too high values for high wind velocities. 



Evaporation at the control station. — The evaporation values obtained 

 with instruments of type 2 in 1916 and of type 4 in 1917 to 1921 are 

 not closely comparable because of differences in shape and in the size 

 of the evaporating areas. They are given together in Table 20, 

 however, because the comparative data for two stations are in part 

 dependent on this 1916 record from the type 2 instrument. 



Although the evaporation record of Type 4 instruments has been 

 secured only for parts of four years, and the averages by decades and 

 months can not be considered even as close approximations to ''nor- 

 mals," still this record is considered very valuable in several respects, 

 viz: 



(1) It shows the magnitude of the evaporation factor during the 

 winter months, due primarily to the somewhat warm and dry char- 

 acter of the descending anticyclonic winds. As already suggested, 

 these winds give the Pikes Peak region a peculiar character, and an 

 examination of the other evaporation records shows that no other 

 region studied has winter conditions so severe. 



