On the Lunar Cloud-Period. 



315 



from January 1, 1900, to January 13, 1907. The results are given 

 at length, in the large Table at the end, in four periods : — 



I. 21 lunations, 1/1/00-13/9/01 ; 

 II. 21 „ 14/9/01-26/5/03; 



III. 21 „ 27/5/03-25/1/05 ; 



IV. 24 ,, 26/1/05-13/1/07. 



In each of these the average percentage of cloud is given at 2 p.m., 

 8 p.m., the mean of six obversations a day, i.e. — 



VIII. + XI. + XIV. + XVII. + XX. + XXIII. 



and the decrease of cloud between 5 p.m. and 11 p.m., for each day 

 of the lunar month. This last, the decrease, is evidently a necessary 

 process, because although we should expect to see more clouds at 

 night in full moonlight than at other times, yet if there be really a 

 cloud-dispersing effect, the difference would be greater about full 

 moon than at other times, other things being equal. In arranging 

 the results the cloudiness has been tabulated for each day of full 

 moon, and for fourteen days before and fourteen after. The thirtieth 

 day, the averages for which are given in brackets, has been assumed 

 equal to the arithmetic mean of the fourteenth day before and the 

 fourteenth after full moon." This day is marked with the conven- 

 tional black dot (•) signifying new moon ; the ring dot (O) signi- 

 fying full moon. The following are the Kenilworth averages in 

 trihemera : — 



Day. 



2 p.m. 



8 p.m. 



Mean. 



Decrease : 

 XVII.-XXIII. 



• -3 



30% 



19% 



24% 



14% 



4-6 



37 



29 



30 



14 



7-9 



40 



31 



32 



12 



10-12 



37 



31 



32 



7 



13-15 



33 



27 



28 



2 



0-18 



36 



33 



32 



7 



19-21 



36 



25 



30 



5 



22-24 



35 



23 



30 



12 



25-27 



39 



23 



29 



14 



28-30 



33 



17 



24 



14 



* Done in 

 29-53 days. 



this way because the lunation, or synodic month, averages only 



