752 J. A. BROUN ON THE LUNAR DIURNAL VARIATION OF 



combining the lunar diurnal variations for each of these phases with those for 

 three days before and three days after them. 



44. The curves for the new moon have been projected in thin lines, and 

 those for full moon in thick lines on the same zero line (see Plate XXVII.) ; 

 similarly for the curves for the first and third quarters in thin and thick lines 

 respectively. In each case the night part of the curve is distinguished from 

 the day part, the former being a dotted, the latter a continuous, line. 



45. The principal conclusions to be drawn from these curves are the follow- 

 ing :— 



1st, The great difference of the amount of movement during the day and 

 during the night is shown distinctly in each case. 



2d, If we consider, in the first instance, the curves for the new moon, we 

 find in the months of January to May, that the north end of the needle changes 

 its direction of motion, and begins to move rapidly eastward at sunrise (almost 

 exactly at sunrise). This holds true till new moon in May ; but in June the 

 direction is completely reversed, the north end of the needle turning westwards 

 shortly after sunrise ; this continues in July, August, and September, but at the 

 new moon in October the reversal again occurs more distinctly than in May : 

 the needle now moves rapidly eastward after sunrise till June again. 



3rd, The same statement holds for full moon, only that the movements in 

 May and October are combinations of those for the preceding and succeed- 

 ing months. I have no doubt that the change would be shown equally well and 

 equally rapidly from one direction to the other, if the full and new moon always 

 happened at exactly the same points of the earth's orbit, but these results are 

 derived from combinations of observation for full moons occurring in the begin- 

 ning and in the end of May and October, and therefore partaking partly of the 

 character due to the preceding and following months. 



4:th, When we examine the curves for the first and third quarters, which are 

 projected together, we find for both quite similar results to the preceding, only 

 that the motion after sunrise is towards the west in January and towards the 

 east in June. 



46. Thus, if sunrise happens when the moon is near the meridian of 18 h. 

 (that is, at new moon, the sun rising always near 18 h. at Trevandrum), or 

 near the meridian of 6 h. (full moon), the needle then turns from its previous 

 direction and moves rapidly (relatively speaking) eastwards in the months from 

 October to May, and westwards from May to October. When the moon is on 

 the meridian of 12 h., or h. at sunrise, the directions of movement are the 

 reverse of the preceding. That is to say, whether a maximum or minimum of 

 easterly declination should happen at sunrise, the movement following takes 

 place comparatively rapidly. 



47. If we had found previously that a minimum or a maximum always hap- 



