Barometric Pressure. 37 



wave of pressure, another fact worth attention is the relative 

 magnitude of the amplitude of the second harmonic term of 

 temperature variability. Relatively to the first harmonic term it is 

 twice as important as the corresponding term in the formula for 

 temperature variation. 



[Note. — Since this address was delivered Mr. Innes has called 

 my attention to a remarkable paper by Dr. Halm, published in the 

 Journal of the Scottish Meteorological Society, Third Series, No. 

 xxii., 1906. Dr. Halm traces the diurnal variation of the barometer, 

 not to the temperature variability, but to the " change of the change " 

 of diurnal temperature, i.e., to d"^/dt", where 6 is the absolute 

 temperature at the time t. I should have been glad to have included 

 a sketch of Halm's results had I known of them before. At the 

 same time it seems to me that the temperature variability is more 

 important than its rate of change.] 



NOTES ON THE TABLES. 



Daressalam. — I have taken the monthly means of Barometric 

 Pressure from Deutsche Uebersecische Met. Beohaclitungen, H. 

 XIII., 1905, and computed from them the annual means and 

 periodic constants for the two years 1901-2. The barometric 

 observations made before 1901 seem to be seriously in error. 



KiMBERLEY. — My own observations. Monthly means of the hourly 

 values of Barometric Pressure are given in Table 4 ; the diurnal 

 variation of Air Temperature and Relative Humidity in Tables 

 5 and 6 ; harmonic constants of Barometric Pressure, Tempera- 

 ture Variability and Humidity in Tables 9 and 10. 



Greenwich. — The Barometric Pressures are taken from the Grcen- 

 zvich Observations, 1883-1894. Monthly means are given in 

 Table 4A. The harmonic constants for the individual months 

 are copied from Harmonic Analysis of Hourly Observations, 

 Etc., at British Observatories, p. 20, 1891 ; but the constants 

 for the year were computed by me from the annual values of 

 Table i here. A long series of observations seems 



required to give good monthly values of the harmonic 

 constants at Greenwich. The harmonic constants of Temperature 

 Variability are given in Table 8. They were computed from 

 the hourly differences deduced from the hourly values _ of Air 

 Temperature given in Harmonic Analysis, Etc., p. 10. The 

 Noon values of Air Temperature at Greenwich seem to be too 

 great throughout the summer. 



