It is shown that the dynamical pressure changes may be replaced by upper pressure waves, under some special assumptions, thus the differential equation for pressure change due to pressure and temperature wave is obtained. Afterr solving this equation under different initial conditions, it is seen that, under a special condition, the resonance of pressure with temperature waves may occur. In the last section, a typical example is given and the figures for thechanges of pressure distribution are obtained.
Using Shibagaki's method of numerical integration, the collection efficiency (the proportion of captured particles) of a circular cylinder is determined theoretically with sufficient accuracy for practical purpose. The obtained values differ considerably from those obtained by F. Albrecht, but nearly coincide with the approximate values computed by the author 9 years ago. The results obtained are applied to the problem of icing on the electric cable, and it is suggested that there is a certain limit to the amount of ice, which will grow in one winter season. The breadth and the density of capture along the surface of a cylinder is also discussed. Finally it is shown that the deviation from Stokes' law arises, when a fog particle approches the cylinder.
Improvement of the nucleus-counter of Aitken type was done to use it in routine works.The principal points of impyovement are as follows: (1) A rubber sheet was used at the pump, just as in Wilson's cloud chamber, to prevent the leakage of air. (2) For counting drops, a microscope was used instead of a lens. It made counting easier to all observers, and moreover we could count drops, with another eye-piece of different power of magnification, to a larger extent of the number of drops without using a dilution-pump. (3) As a dilution-pump, Scholz's type was adopted. We could measure necessary volume of air more accurately by this pump than by Aitken's one. (4) No leakage of air occurred through the pump in each stage of successive expansions of the same sample of air, thus it enabled us to measure neuclei-spectrums using various ratio of expansions. (5) According to the result of investigations of F. J. Scrase, the stirring-vane was removed from this chamber. (6) This apparatus was tested at various heights of Mt. Fuji both in summer and in winter seasons, and proved its excellent functions.
It was found out that a thin boundary layer in which no fog is formed exists along the wall surface of the cloud chamber, when dusty air is taken into the chamber and expanded with expansion-ratio of about 1.2. The mean thickness of this boundary layer, which was measured with a microscope in various ways (see figure), was 0.25mm. The thickness seemed to be related to the speed of the expansion. The above figures were observed when the speed of expansion was about 0.03 sec. When the speed was slow as in the case of nucleus counting, the thickness increased becoming 0.4_??_0.8mm. The formation of this layer seemed to be explained by the thermal conduction from the wall and also by the decrease of density of water vapour in the chamber due to expansion. The above phenomena, the wall-effect, have a great important role in the theory of m_??_asurement of nuclei. Due to the existence of such a layer, it is true that nuclei fallen on a glass plate are not the entire nuclei that exist in the chamber, i.e. nuclei included in the layers near the glass plates may be left unchanged into drops. Thus we have to repeat expansions successively on the same sample of air in order to find out the total number of nuclei in the chamber. Assuming the thickness of the layer, the number of drops that falls at each stage of expansion was theoretically calculated and the result was compared with the observed values obtained by J. J. Nolan, G. R. Wait and other workers. It was proved that number of drops caught in each expansion seemed to be explained mostly by the existence of the layer.
Fluctuations of air temperature in the air layer up to 15 meters above the ground were measured by resistance thermometers (which were made of bare nickel wire, 0.03mm in diameter and about 80cm long, stretched in the space without protecting cover). This thermometer was inserted in one arm of a Wheatstone bridge and the unbalanced current was lead to an electromagnetic oscillograph. The errors that occur in the measurement are discussed. They consist of those induced by the wind pressure, insolation and time lag of the thermometer. Their magnitudes are generally less than 0.04°C, on the other hand fluctuations of temperature are about 0.4°C. So, errors described above are negligible in the measurement of temperature fluctuation. The magnitude of fluctuations of air temperature is 0.3_??_0.4°C in the air layer 1 meter above the ground and the magnitude decreases to 0.2_??_0.1°C in the air layer 15 meters above the ground. Timely and spacial correlations of air temperature fluctuations (Rξ and Ry) are obtained. Rξ is expressed approximately by the exponential curve e-μξ, and the magnitude of is about 8_??_16 sec. (where T is the value ofξ at which Rξ first cuts the ξ-axis.) Ry is positive even at y=14 meters. The values of correlation are larger than usual when the air temperature of the ground is increasing rapidly. It is discussed that fluctuations of temperature are not wholly due to the mixing phenomena (i.e. the random motion of the lump of air).
The time variations of the amount of nitrate, nitrite and ammonia contained in thunder-shower were measured by using Griess Romijn's reagent for nitrite and Nessler's reagent for ammonia. The author supposes that the ammonia content must be the product of the daily living of people from the fact that the amount of ammonia decreases to almost zero in about three hours. Nitrate is supposed to be the product due to the thunder-discharge based upon the fact that the amount of nitrate does not vary with the lapse of time, and also that the amount of nitrate increases rapidly after the occurrence of violent thunder. Lastly, the nitrite content was also concluded to be the product of thunder-discharge. The mean values of the amount of nitrogen per liter of shower water in the states of nitrate, nitrite, and ammonia were as follows: 4.3γN/L…nitrogen in nitrite state. 0.197Nmg/L…nitrogen nitrogen in nitrate state. 0.346Nmg/L…nitrogen in ammonium state. The time variation of the amount measured is shown in Fig. 1.