THE INTERIOR OP THE EARTH. 253 



to compute at least the order of the effects in the latitude variations 

 that we should expect from these various causes. Furthermore, the 

 annual term is probably variable in its amplitude, and it is important 

 to ascertain how (if at all) these changes are related to the corre- 

 sponding meteorological observations. 



One other term must be mentioned in this brief summary. A few 

 years ago Kimura of Japan made the important discovery (the 

 most striking contribution to astronomy that has ever come out of 

 Asia) that tlie latitudes of all stations are affected by a variation 

 that does not depend upon the longitude but which is the same for 

 all points in the same latitude. In other words, there is present a 

 variation that is not due to the wanderings of the pole. To ascertain 

 more closely the nature of this term, the International Geodetic Asso- 

 ciation extended its latitude service temporarily to the Southern 

 Hemisphere, with the result that the term was found to be of pre- 

 cisely the kind that would be caused by an annual wandering of the 

 center of gravity of the earth to and fro along the axis of rotation. 

 This must be regarded merely as an illustration and not as an ex- 

 planation, for so great a change (about 3 meters) in the position of 

 the center of gi-avity is excluded on other and very conclusive 

 grounds. No plausible explanation for the Kimura term has as 

 yet made its appearance, and as a consequence the reality of the term 

 has been questioned from every possible point of view. Many ex- 

 planations have been advanced, each of which sought to account for 

 the term as merely an instrumental effect or the like, just as v>^as the 

 case 20 years earlier with the whole of the latitude variation itself. 

 Against such attempts the Kimura term has held up very well. It 

 is not too much to say that at the present time all but one of the 

 numerous explanations of this class have been disposed of; this 

 exception deserves a brief mention, particularly as it calls loudly for 

 the attention of the meteorologist. Let us suppose that the layers of 

 equal density in the atmosphere above a station are not horizontal, 

 but that they are sensibly inclined. If this occurs without our 

 laiowledge, as it would under ordinary circumstances, then w^e 

 should apply refraction to our observations in a slightly erroneous 

 way and we should derive a value for the latitude that is not quite 

 correct. Let us suppose further that this effect were a world-wide 

 one and that in any given month there would be a pronounced ten- 

 dency for the inclination to be in the same sense in all latitudes, north 

 and south, as well as in all longitudes. Then we should have a set 

 of circumstances that would account for the Kimura term as an 

 atmospheric effect, and therefore it would be excluded as a real varia- 

 tion of latitude. So far as the astronomer is able to testify, the evi- 

 dence is against the occurrence of such tilts in the atmosphere. The 



