316 



ENVIRONMENTAL CONDITIONS. 



rainfall and of relatively low evaporation values. This drought- 

 breaking rain occurred in the week ending July 20 at Tucson, in the 

 next following week at Easton. 



Miami, Florida (fig. 12) : The exceptionally long record for Miami 

 shows a maximum evaporation rate of 232 c.c. early in the season 

 (week ending May 18), and a minimum of 35 c.c. near the end of the 

 series (week ending October 5). While the general trend of the graph 

 of atmospheric evaporating power is downward as the season advances, 

 there are rather pronounced variations, which have some tendency to 

 occur periodically, with an interval of about 8 weeks. The intensity 

 of precipitation appears to increase throughout the season, and heavy 

 rains have relatively but little relation to the evaporation values. 



JOB 



261 



203 



IS6^ 158 



na 



§L 



133 



83 



128 



Hi. 



>2Z 



I3S 



lAZ 



9S 



EvnpannTiOM 



"3 112 



S9 



98jl06 



120 



63 



102 



Hi. 



/2S 



S9 



A2i 



392 



2.93 



2.6^ 



l.-fe 



/.-H 



1. 35 



1.36 



160 



FniNFflLL 



.13 



.36 



33 



.S3 



■S9 



.62 



dZ. 



.0-* 



JZ 



34 



OO 



.02 



jOS 



ae 



f22. 



jOO 



/}0 



I Mftf'^ II IS asjuMfi B IS az isJuire ij Ta eiflusa To n 2-* 3isej> n ^* aiesOcrs iz is eeim 

 Fig. 11. — Weekly precipitation and evaporation indices, summer of 1908, Easton, Maryland. 



Mean evaporation values for 5-week periods and for 15-week season 

 (Table 17, plate 56.) — As has been indicated, the average weekly rates 

 of loss from the porous-cup atmometer, at the several stations, have 

 been calculated for the 5-week periods, April 20 to May 25, May 26 to 

 June 29, June 20 to August 3, August 4 to September 7, and September 

 8 to October 12. For a few stations the average rates for the last 3 

 weeks of the series (from October 12 to November 2) have also been 

 calculated. These averages are given in the final columns of table 16 

 and they are all brought together in table 17. Where data are not 

 available for all the 5 (or 3) weeks of a period, the average has been 

 made from the smaller number of weekly records at hand, and the 



