48 



plot and plots 10 and 11, show an even heavier loss due to 

 rainfall in July and August. 



Rainfall variations make an important contribution to the 

 yield variation observed. In this respect rain is perhaps more 

 important than any other single meteorological factor. It will 

 not be possible to treat the other meteorological factors with the 

 same precision, since the records of temperature and sunshine do 

 not go back to the beginning of the experiments. 



See also paper No. XVII. 



II. METEOROLOGY. 



(Physical and Statistical Departments.) 



XIII. W. B. Haines. " A Comparison of the Radiation 

 Recorders at Rothamsted. Journal of the Royal 

 Meteorological Society, 1925. Vol. LI., pp. 95-100. 



This paper deals with a comparison of the readings taken 

 at Rothamsted with three types of radiation recorder. The first 

 is a recorder of the Callendar pattern, depending upon the differ- 

 ence in temperature between a black and a bright resistance 

 exposed to the sky. These readings are taken as standard. The 

 second instrument (the Wilson Radio-integrator) reads the 

 amount of alcohol which distils from a bulb exposed to the 

 radiation into a similar shielded bulb. The third set of data 

 is the record of hours of bright sunshine from the widely used 

 Campbell-Stokes apparatus. Reference is also made to a fourth 

 set of data, that given by an evaporimeter of the porous candle 

 type, since the readings of this instrument are correlated to the 

 amount of radiation. 



The alcohol integrator gives readings much too low during 

 the winter months. The readings can be fitted with fair accuracy 

 by a formula of simple parabolic type. The possibility of intro- 

 ducing a temperature correction is discussed. 



The hours of bright sunshine should be corrected by a factor 

 depending upon the time of day and year (i.e., upon the sun's 



o 



altitude). A formula deduced by Angstrom from the Stockholm 

 data, for calculating total radiation from hours of bright sun- 

 shine, is examined and found fairly satisfactory for the Rotham- 

 sted data. It is concluded that such a formula, based upon the 

 data at one station, could with due caution be adopted for another 

 station. 



The evaporimeter results follow the hours of sunshine very 

 closely, but some care is needed in the choice of a site for this 

 instrument. 



XIV. W. D. Christmas. " Notes on the Weather at 

 Rothamsted/' " Nature," Oct. 27th, 1921 ; Jan. 16th, 

 1922. " The Times," Jan. 26th, Julv 4th, Aug. 2nd, 

 Sep!. 3rd, Oct. 1st, Nov. 2nd, Dec. 2nd, 1923; Jan. 

 2nd, Mar. 1st, June 2nd, Sept. 1st, 1924; Jan. 1st, 

 L926. 



