coo Journal of Agricultural Research voi. xvni, no. io 



or daytime period, has so many physical factors involved that the 

 problem was very difficult. The night interval, or lower part of the curve, 

 they found to be logarithmic, the constants of the equation having a 

 physical significance. The equation of the upper part of the curve, the 

 equation of the lower part of the curve, and a combined equation of the 

 two — partly empirical — have been worked out. Occasionally a section 

 of the country experiences a warm winter or a cold summer.^ These 

 temperature courses have been studied by Gawthrop.* 



ANNUAL TEMPERATURE CHANGE 



The seasonal change in temperature is due to the fact that the earth 

 is moving around the sun in an elHptical orbit with a variable speed, its 

 axis of rotation making an angle of 23^° with the plane in which it 

 revolves. The 24-hour temperature variation is due to the rotation of 

 the earth on its axis. 



If the rate of emission of heat by the sun were to remain constant and 

 if the rate of rotation and of translation of the earth be the same each 

 year — the path of the earth around the sun being the same each year — 

 and if the diathermancy of the atmosphere were to remain constant with 

 no evaporation or condensation, the sequence of temperatures of one 

 year would be very nearly exactly that of any other year, and the therm- 

 ograph record, for April I, for example, would be the same every year. 

 On cloudless days this is nearly realized. The foregoing conditions are 

 nearly realized except for the diathermancy of the atmosphere, which 

 varies because of evaporation and cloud formation. It is because of 

 the passage of cyclones and anticyclones — storm and fair weather areas — 

 across a section with the accompanying rain or snow, the dissolving of 

 clouds, or evaporation of rain, that we have temperature departures 

 from normal. 



In order to determine the normal change in temperature with the time 

 over the entire year, it would be necessary to get the mean temperature 

 for every hour of the year for enough years to eliminate the irregulari- 

 ties due to storms, and then make a plot of these 8,760 (24 X 365) hourly 

 temperatures and determine the equation of the curve. This is possible, 

 but it is obviously a very tedious operation. The same result may be ob- 

 tained much more easily by dividing the problem and considering the 

 seasonal temperature change and the daily change separately. In other 

 words, it will be shown how to determine the mean daily temperature, 

 and this value will be multiplied by a certain percentage in order to get 

 the value for a certain hour of the day. 



However, the arid West, comprising the section between the Rocky 

 and Sierra Nevada Mountains, has only about 10 inches of rainfall and 

 little cloudiness, a humidity of but 50 per cent, and 300 days of the year 



' Pbrnter, J. M. OP. at. 



'Gawthrop, Henry, temperature courses. In Mo. WeatUer Rev., v. 3s, no. 12, p. 576-378, i fig- 1907- 



