The empirical fact the systematic night-errors of the latitude observations appear to depend mainly on the direction and velocity of the wind, that is, the wind effect has been investigated for many years at the Royal Observatory, Greenwich and the International Latitude Observatory, Mizusawa. The principal aim of this note consists in investigating the probable relation between the anomalous refraction due to the tilting of the air-strata of equal density and the wind effect, The pressure gradients, the temperature gradients and the inclinations of the air-strata of equal density in the upper air were deduced from the wind velocities. The inclinations of the air-strata of equal density are north up as far as about 2, 000 m through the year. Teey correspond to the positive, residal latitudes producedd by the wind, as expected theoretically from the relation between the anomalous refraction and the observed latitudes.
The annual variations of these inclinations give their maxima in winter and their minima in summer, showing the remarkable similarity to those of the wind effect in the same interval. Considering several other characteristics in common with each other, the anomalous refraction due to the tilting of the air-strata of equal density might be a main, physical cause of the wind effect on the latitude observations.

抄録全体を表示

Jåderin wires being very fine and apt to accept dammages from various handlings it has been suggested to standardize them in the field of survey immediately before and after the work. For this purpose, the use of light wave standard is not only suitable for its practice, but also appropriate to avoid many systematic errors which may happen to be introduced in the usual micrometric method based on the prototype meter . In Japan, however, where the climate is very unstable, it has been very difficult to measure long distance by interference of light waves. The measurements have been realized by underground or similar installations. We have made experiments to find the possibility to execute the measurement in the open field [20] and arrived at an idea of an interference comparator which may be applied to practical use. In the present report, are described the circumstances how to overcome the instabilty of the atmospheric conditions and the mechanical disturbances and the preliminary design of the apparatus (Figs 13 and 14). At the same time, a supporter of Jåderin wire is designed by use of three knife-edges (Figs 17 and 18) to avoid friction at the supension point of the wire. it is found easy to guaranty the weight of ±0.5g. to 10 kg. by this supporter.

抄録全体を表示

A method for calculating subterranean mass distribution etc, from the gravity value at the surface is discussed for the three dimensional case. The principle of the method was already described in the preceding papers.
In the three dimensional case problems newly arise which are relate to the upper boundary of the spectrum of the gravity data. That is, different unit function or unit gravity must be used according to the assumed spectrum boundary.
In the present paper three kinds of unit functions are proposed, and these are summarized as followings.
1) When the upper limit of spatial wave number of the gravity data is independently given for x and y direction the unit function represented by
1)
is used.
2) When the upper limit of wave number is given in such a way as √ω₁²+ω₂²<Ω, the unit function represented by
is used,
2) where ω₁, ω₂ respectively represents wave number of x and y direction.
3) When the limit of wave number is limited in the hexagonal region on the wave number plane, the unit function represented by
is used.
The unit function δ_??_ is product form of trle unit function aiscussea in me two aimensional case. The gravity value given at lattice points is replaced by the smoothed function whose value concides with given value at each lattice points. The operations are weighted sum of that unit function. The mass distribution etc, are given by the weighted sum of the response for that unit gravity. A week point of this unit function is that the assumed limit of wave number varies according to the choice of the co-ordinate.
As it is natural that the wave number in the two dimensional spectrum is given by √ω₁² the unit function δ_??_ is reasonably used. This unit function is regorous but contains inconyenience, if we only consider simplicity of calculations. That is, the smoothed function which takes the place of the given data is given by the weighted integral of continuous data. And the mass distribution etc. are also given by the weighted integral of the response for that unit function. As a complomise of 1) and 2), it is desirable that both the upper limit of wave number is given as the case 2) and the results are obtained by the weighted sum of the data. In order to satisfy these conditions as much as possible, a hexagonal unit function δ_??_ is proposed. If this unit function is adopted, and if the data are given at the mesh points which are represented by the intersection points of the groups of straight lines having equal distance and each group are mutually inclined by 60°, the smoothed function and the necessary results are obtained as weighted sum of the unit function or of response for that unit function.

抄録全体を表示