Studies in Prairie and Woodland 21 



Readings were made and the water replenished from time to 

 time according to the best technique which has now become so 

 nearly standardized as to require no detailed description here. 

 Suffice it to say that the data here considered are all from stand- 

 ardized rain-correcting instruments to which the proper correc- 

 tions have been applied. The porous cups were run for periods 

 of 5-6 weeks after which they were replaced by newly stand- 

 ardized cups and the whole apparatus cleaned and started afresh. 



The graphs on evaporation have been drawn by first computing 

 the average water loss in cubic centimeters per day between the 

 successive readings and then these data are plotted as the ordi- 

 nates while the intervals between the readings are indicated by 

 the abscissae. This method facilitates the comparison of evap- 

 oration quantities from the various stations and it also depicts in 

 a graphic manner the march of evaporation for each of the re- 

 spective stations throughout the season. 



Some striking facts are disclosed by a comparison of the 1917 

 evaporation graphs with those for the 1916 season. In the first 

 place it is seen that the season of 1917 was marked by a very 

 much higher evaporation rate which began in early spring and 

 continued practically throughout the summer. The maximum 

 evaporation in the prairie station in 1916 came during the latter 

 portion of July and the fore part of August when the extreme of 

 43 c.c. was reached. The next highest figure was late in May 

 when the evaporation was slightly above 25 c.c. In comparison 

 with these figures we see that the highest maximum for the season 

 of 1917 was 47 c.c. (Fig. 10) and this occurred the latter part of 

 April. But there were also three other secondary maxima, one of 

 45 c.c. during May, another in the middle of June of 39 c.c. and 

 another near the close of July of 35 c.c. All of these are high as 

 compared with the evaporation rates for the season of 1916. 

 Correspondingly higher evaporation rates were also recorded for 

 the thicket and woodland stations in 191 7 as is readily noted by 

 comparing the march of evaporation in those two stations as de- 

 picted in the two sets of graphs. 



Our data concerning the evaporating power of the air thus in- 

 dicate that aerial conditions in all of our stations were very much 



27 



