This is an application of theories of the variability of natural phenomena which were established by Mr. S. Ono. (On the magnetic characterization of days and the variabiliy of natural phenomena; Annual Report of the Central Met. Obs. of Japan, Magnetic Observations for the Year 1916, published in 1924, Japanese Journal of Astronomy and Geophysics Vol. III, No.1, 1925, an abstract; Terrestrial Magnetism and Atmospheric Electricity, March 1924, an abstract in Japanese; this Journal Ser. 2, Vol. 3, No.5, 1925. Investigation on range of the variation of magnetic elements; Annual Report of the Central Met. Obs. of Japan, Magnetic Observations for years 1917 and 1920, an abstract in Japanese, this Journal Ser. 2. Vol. 3 No.8) Present author tried to apply these theories on the variation of temperature. He used thermograms observed at Tokyo between April 1914 and January 1915. To specify four seasons, A pril, July, October and January are selected, and in each of these months seven representative days are used in which two most disturbed days, two days without disturbances except smooth diurnal variation and three moderately disturbed days are contained. The temperature was read on curves for every 15 minutes and difference of contaive two are designated by a simbol ξ. Besides this quantity the daily range Δ is observed for each day.
Numerical results to compare _??_ (the mesn of ξ), (_??_)
2 (the square _??_), _??_ (the mean of square of ξ) and Δ are shown in a table which follows conalusions: Though there is an ample range of magnitude of _??_, _??_ or Δ, relations prived by Mr. Ono _??_=2(_??_)
2 and _??_/_??_=1/2U hold good in cages of temperature variation. In this equation _??_ is the mean of Δ and U is the time duration without disturbance in a day and the square of the mean of daily range can be used to express the monthly or the seasonal mean variability of temperature variations. Value of 1/2U in present tesult is very near to 48. Since the unit of time for ξ is taken 15 min., U becomes 2×48×15/60 hours
i.e. 24 hours. This result means that the mean duration of disturbances is negligibly short compared with a day in the case of the temperature variation.
View full abstract