Many papers have been published which discuss Bouguer anomalies, from the viewpoint that they give equivalent anomalies "ΔgD" at a certain depth and that they therefo e give important informations about the underground construction of the earth. In these discussions, almost always Bouguer anomaly itself was used. The spectrum of Bouguer anomalies which has the same information, may more conveniently be used, and by doing so, the spectrum of equivalent underground "ΔgD" can be obtained more easily and will gives us the same information about the underground equivalent " ΔgD". If we take suchstand point of view, the first problem will be to see, whether the spectrum of underground equivalent "ΔgD" does converge or not. If we are to obtain equivalent "ΔgD" at a certain assumed depth, it is necessary that the spectrum of "Δg" at that depth is convergent. The spectrum must become gradually small according as the degree of the spectrum increases. In this paper, the spectrum of surface "Δg" was deduced for central and southern part of Japan. The spectrum intensity corresponding to these areas are approximately represented by e^-ωD. This means that the spectrum of "ΔgD" becomes approxi mately white at a depth D. Thus we can concider D as the limit of the depth where the equivalent "ΔgD" is reasonably accepted. This limit of depth has been obtained for many section in that area, and is shown in the Fig. 7. In the case of section No. 0, from the results of the explosion studies, we know the depth of the Mohorovicic layer. The limit of the depth as determined by the above method coincides well with the result of the explosion studies. This means that at this depth the spectrum of the equivalent "dg" is white. That is, the undulation of the interface of the Mohorovicic layer has a white spe ctrum.