394 



the total loss of sunsliiue in the year to bo 102 hours at Fort Collins 

 and 93 hours at San Luis. A tabular record of the amount of sunshine 

 at three places in Colorado is given for each month of 1889. 



Solar radiation.— The intensity of the sunlight as measured by the 

 solar radiation thermometers at the station has been very great. In 

 February the radiation thermometer registered 117° above the ther- 

 mometer in the shade close by. " On the average, the excess is about 

 (jQo * * * Ttie terrestrial radiation thermometer will average about 

 7° less than the air thermometer in the instrument house, G feet higher. 

 The diflerence rarely exceeds 15°, though 24° has been registered dur- 

 ing the year." 



Evaporation. — " Evaporation, in view of the questions arising from 

 water storage, is of considerable importance. Measurements were 

 made on tanks placed in the ground, here and at the substations, and 

 also on tanks floating in the water of reservoirs, canals, and the river. 

 The tanks used for comparison here and at the substations were 3 

 feet square and 3 feet deep. At this station smaller tanks, to deter- 

 mine the influence of size and of the material composing the tank, were 

 also used. Measurements were made with the hook gauge at 12-hour 

 intervals, during June, July, and August, and once per day in Septem- 

 ber and October. 



"Evaporation depends on a variety of conditions. The amount 

 fluctuates greatly from day to day. Nevertheless, we succeeded in 

 producing a formula for computing the evaporation, which agrees closely 

 with the observed values. The evaporation, as computed for the 5 

 months from June to November, difters from the observed evaporation 

 by only a quarter of an inch. Our measurements show that the 

 evaporation, as measured on days of storms, is frequently excessive, and 

 consequently measurements made at long intervals give too great an. 

 evaporation. The evaporation for the whole year, from our principal 

 tank, is less than 40 inches. At the other stations it has been much 

 greater." 



A table gives the monthly evaporation for 3 years (1887-89) at the 

 station, and for 1889 at the two substations. 



The observations made oa one tank from June 1 to November 1 were submitted to 

 a reductiou, and tlie formula found -with the most probable co-efficients, as deter- 

 mined by the method of least squares, is E — .1934 (T-t) (1 + .005 W) where E is the 

 evaporation iu inches for 12 hours ; T is the vapor tension at the temperature of the 

 •water surface; t, the vapor tension of the air; and W, the movement of the wind 

 in miles during the 12 hours. This formula satisfies our observations very closely. 

 The computed evaporation for the 5 months from June 1 to November 1 differs only 

 a quarter of an inch from that observed. 



For the evaporation of a whole day the formula becomes, E = .3868 (T-t) (1 -f .0025 

 W). Tbe formula found by Fitzgerald, in his careful experiments at Boston, 

 expressed for 24 hours, would be : evaporation for 1 day = .3984 (T-t) (1 + .0208 W). 

 The close agreement of these co-efficients from investigations carried on under as dif- 

 ferent circumstances as these were, strengthens the confidence in either formula, and 

 makes it probable that the true value of the co-efficient is not far from 0.39 or 0.40. 



