AND ITS CHANGE WITH LUNAR DISTANCE. 175 



NOTE ADDED MARCH, 1915. 



(l) The Seasonal Changes of 6> P A . 



If the values of 6 P A from Tables III. (a), (b) and IV. (a), (b) be grouped together 

 according to their season, it will be found that the mean value at the equinoxes is 

 about double that at the other seasons, the difference being quite noticeable ; this 

 was kindly pointed out to me by Prof. H. H. TURNER, F.Il.S. In order to examine 

 this point a little further (since the result mentioned rests on material which is much 

 more scanty for the equinox than for the other seasons), I caused the results given in 

 Tables I. to IV. (c) (for vertical force), to be computed. These were not in the 

 paper as originally communicated to the Royal Society, as I did not then know that 

 the amplitudes of the variations for this element and these seasons reached the 

 value 2y. 



If now all the values of 6 P A in Tables III. and IV. are grouped together 

 according to season, and the means taken, we obtain the following results : 



Mean Values of 9 P A from Table III. (short periods). 



Degrees. / 



Summer 21 + 3'2 from seven determinations. 



Winter 26 2' 7 five ,, 



Equinox 30 + G'G ,, five ,, 



Mean Values of L , A from Table IV. (long periods). 



Degrees. 



Summer* G'50'9 from six determinations. 



Winter 11'21'1 ,, five 



Equinox 14'72'G ,, five 



Both groups of mean values show a progression in the phase change from summer 

 to winter, with maxima at the equinoxes. If this is really the case, the phenomenon 

 should make success in the search for the cause of the phase change much more 

 probable ; at present, however, the magnitude of the probable errors are such as to 

 cast doubt on the reality of the seasonal change, and, until it is more definitely 

 established, any attempt at its explanation would be premature. 



(2) The Lunar Atmospheric Tide. 



Further information concerning this is contained in a memoir by WAGNER,! who 

 has discussed the barometric observations for the years 1903-8 made at Samoa. The 

 mean result for the semi-diurnal tide is, in millimetres, 



0'039 sin (2< + 33 degrees). 



* Omitting the negative value - 10 for Pola horizontal force, 

 t G. WAGNER, 'Gottingen Abh.,' IX., 4 (1913). 



