588 



Prof. B. Stewart on the Daily 



Applying this correction to (D), we obtain the following corrected results 

 for the winter lunations of the whole series : — 



Phase of lunation... (0) (1) (2) (3) (4) (5) (6) (7) 

 Correct value of win- 1 „ /-r^v 

 ter lunar range... I "'^^ "'37 "'S^ io-88 10-52 10-49 10-79 ^^'^S (F) 



Series (F) is represented in Fig. 3. 



11. If we now make use of the lunations corresponding to the six 

 summer months (April to September), emplo>dng for this purpose these 

 lunations, not specified in the winter list already given, we obtain the 

 following result : — 



Phase of lunation... (0) (1) (2) (3) (4) (5) (6) (7) 



Value of range i6-86 16-98 i7'2.5 17-26 17-43 iJ'^i i7'26 17-23 (Gr) 



To these values must be applied a small residual correction of the same 

 nature as that represented by (E), but opposite in sign to it. This cor- 

 rection, obtained in the same manner as (E), is as follows : — 



Phase of lunation... (0) (1) (2) (3) (4) (5) (6) (7) 

 Correction appli- 1 



cable to summer i» -{-'lo +'04 —-02 —-04 —-08 —'06 —-02 +-04 (H) 

 months. J 



Applying this correction to (Gr), we obtain the following corrected re- 

 sults for the summer lunations of the whole series : — 



Phase of lunation ... (0) (1) (2) (3) (4) (5) (6) (7) 



Correct value of sum-") , ^ ,-rx 

 merlunarrange...r^'96 i/'o^ 17-23 ^T^^ ^J'ZS ^T^S ^r^\ ^T^l (I) 



Series (I) is represented in Eig. 2. 



12. If we look at Eig. 1, which represents the lunar variation for the 

 whole year, we shall observe, as already remarked, that it seems capable 

 of being represented by a curve of double period superposed upon one of 

 single period, which latter has its left-hand values greatest, or, in other 

 words, is similar to Eig. 3. If we now look at Eig. 2, which represents the 

 lunar variation for the summer months, we perceive an irregularity as if 

 two variations were superposed on each other, one a double variation and 

 another a single variation of opposite nature to that of Eig. 3. Eig. 3 

 speaks for itself ; the variation is here extremely marked ; and if we sup- 

 pose it due to the superposition of two curves, a double and a single one, 

 we must suppose the single curve to be much more pronounced than the 

 double one. 



13. The hypothesis of a double curve common to both periods of the 

 year, and a single curve going in one way and very marked for the winter 

 months, and in the opposite way and not so marked for the summer 

 months, would appear to be one which would at any rate enable the re- 

 sults to be presented to the eye in a more simple manner. Let us now 

 deduct the variation implied in Eig. 1, or the whole annual variation, 

 from the values of Eig. 3, or the winter values, and we obtain as 

 follows : — 



