The present paper starts with a criticism of the well-known Fowle's empirical formula which gives the amount of solar radiation to be lost by the absorption of the atmospheric water vapour: Fv=0.057+0.012Qm, Q being the precipitable water and m the “air mass” The occurrence of the constant term in the formula, whose absurdity is out of question, is attri buted to the probable perturbation of the occasional existence of the layer consisting of warm moist air of different origin superposed on the air mass in the lower region. This is pointed out on the actual plots of the Fowle's observational data, giving the improved formula F=0.047e (e is the vapour pessure in cm. of Hg., as observed on the earth). The author's opinion is strengthend by the results of observations of total solar radiation made on the summit of Mt. Fuji (3770m.) during the past one year. In the majority of cases, the reduction of the total amount of solar radiation JmT due to the combined effect of the scattered reflection of dust (solid or liquid) and the selective absorption of vapour is remarkably less than that given by the Fowle's formula. This may be regarded a consequence of our being free from the alluded perturbation of the upper air current, as can be easily seen from the synoptic condition pertaining to the great majority of cases. In order to separate those two effects from each other the observations were made through the Schott glass F 4512 (2mm thick). The relative amount of radiation in the visible and U. V. region Sm, and that in the infra-red part Lm were computed by the numerical integrations on the basis of Rayleigh scattering of pure air and the expression of the dust scattering. By means of C-contours in (Sm, m) and (Lm, m) co-ordinates the numerical solution of the equations JmT=const. (Sm+Lm) and JmR=const.K. Lm(aw is the vapour-factor, K factor depending on the transmission of the filter, JmR the radiation amount observed with the red filter) were made to determine C and aw. The result of determination of K-value and the solar constant for the infra-red radiation transmitted by the filter almost exactly coincide with the values computed from the solar spectral energy curve and the transmission curve of the glass, i.eK=0.08989 and solar const.=1.080cal. cm-2 min-1. Moreover the slight dependence of K-value on the air mass is pointed out from the result of numerical integration. The assumed value of x=1.3 was proved to be too small in many cases to give consistent result, which suggests the necessity of increasing it so much as 0.7 or 0.8 in many cases.
In the present paper the authors give the results of observations of the electrical conductivity of the atmosphere at Toyohara for the period extending from August, 1932 to August, 1933, the Second Polar Year. The conductivity was observed with the Gerd_??_en's apparatus, which was made at the workshop of the Central Meteorological Observatory, thrice a day, between 9 and 10 o'clock a. m., at no_??_n, and between 2 and 3 o'clock p. m., The annual variation shows maximum 1.94×10-4 Sec-1 E. S. U. in August and a minimum 0.74×10-4 Sec-1 E. S. U. in April. We can see a comparatively regular daily change from the observations, made on the First and Second International Principal days, twelve times a day. It is noticeable that in winter the conductivity has greater influence cf town than in summer. This may be the consequence of decreased conductivity due to absorption of small ions on the smoke particles. The relation between the conductivity and the wind force or the relative humidity is complicated.
In the previous paper the author introduced a provisory functional expression Φ(s) to represent the effect of prevailing winds on the depth of a day, where s is the length along the coast line, together with an actual example. In Part I of the present paper he gave another example which seems to justify his method of treatment. In Part II lie considered the effect of topography and tried to eliminate it. Thus the cu_??_ve of depth dist_??_ibution Z(s) showed the closer correi_??_tion with that of the wind effect W(s) by taking into account the topographical feature. He also made a discussion on the correlation factor k which combines Z(s) and W(s) by the linear equation Z(s)=K. W(s). The above method of procedure was applied to some actual cases in Part III. The results seem to justify it.