Considerations on the Problem of Latitude Variation. 29 



it is the correction to that term of latitude variation which is in- 

 dependent of polar motion, or z term. 



Now, according to Ross^\ the mean value of z term during 

 the years 1900- 1005, can be put in the following analytical form, — 



+ 0."()27 sin (® + 170°) (® = sun's longitude) 



when tlie effects of stellar parallax, Oppolzer's term in latitude 

 variation and nutation are taken into account and excluded. Hence 

 it follows that the result obtained through my argument is practi- 

 cally sufficient to account for the closing sum and z term. 



Now, as the phenomenon discussed by me is considered to 

 arise from some unsymmetry of the meteorological conditions and 

 also from the solar radiation on the ground, its sense should be 

 inverted for the southern hemisphere and is to be considered as 

 an odd function of latitude, vanishing at the equator. So it can be 

 looked upon as varying with sm f. Under such a conception, the 

 effect on the southern observations should be opposite to tliat 

 on the northern, the correction to the latitude being of positive 

 sign and the correction to the closing sum negative. 



The correction to be applied to the latitude variation indepen- 

 dent of longitude would be 



positively small for January, and 



positively large for July. 

 Therefore, the correction to Ije applied to the yearly term in the 

 variation of latitude, or z term, is 



negative for January, and 



positive for July. 

 The sign is the same as that of the northern observations, and is 

 sufficient to interpret the result obtained from the southern observ- 

 ations. 



As to the amplitude of the z term, it depends jointly on the 

 said phenomenon and the seasonal variation, Ijoth of which can be 

 looked upon as varying with sin <f> ; so we can consider the ampli- 

 tude of the correction of z term as dependent on sl/i^ <p. This 

 signifies that the amplitude of the z term increases from the value 



1) Astronomische Nachrichten, Xr. 4593. 



