is:4.] 



201 



[Chase. 



Jun.— Nov., Jul. — Oct., Sep.— Oct., Sep.— Nov.; 8, Aug.— Oct.; 7, Jan. 

 — Nov., May — Sep., Aug.— Dec; 6, Feb.— Mar. 



The greatest amount of change produced by the lateral smoothing is 

 shown in the following summary of comparisons between Table III. and 

 Table V. : 



Table V. is formed from Table IV. by taking two additional successive 

 means. I am inclined to think that its normals would best represent the 

 means of observations extending over indefinite long periods, but Table 

 III. would perhaps more nearly indicate the disturbances of mean lunar 

 influence that might be expected at different seasons of the year. It is 

 possible that by systematically comparing monthly observations with each 

 of the tables, probable causes for any marked deviations from the nor- 

 mals might be found. 



Table I. presents three sets of solar and six sets of lunar normals, each 

 of which is derived from observations extending over equal, but non- 

 correspondent, periods of one hundred and eight months. They there- 

 fore furnish data for three entirely independent solar, as well as for 

 seven entirely, and three nearly independent lunar comparisons. The 

 lunar columns cover twenty-seven years' observations in the following 

 months : Summer Solstice, May to August, inclusive ; Winter Solstice, 

 November to February, inclusive ; Vernal Equinox, February to May, 

 inclusive ; Autumnal Equinox, August to November, inclusive ; Vernal 

 and Autumnal Equinox, March, April, September, October ; Summer 

 and Winter Solstice, June, July, December, January. The solar columns 

 exhibit, as we might expect, the closest accordance. The lunar, in spite 

 of the great irregularities in Spring and Fall, also exhibit a predominance 

 of accordances in each of the ten comparisons, whereas, if there were no 

 well-marked lunar action, we ought to have found a predominance of 

 disagreements in five of the comparisons. 



The accompanying curves illustrate some of the more important results 

 of the foregoing discussion : 

 Curves 1-12 (Lunar), illustrating Table IV. 



1. January. 4. April. 7. July. 10. October. 



2. February. 5. May. 8. August. 11. November. 



3. March. 6. June. 9. September. 12. December. 

 Curves 13-15 (Lunar), illustrating Table I. 



13. Summer Solstitial, continuous line. 

 Winter ' ' broken line. 



14. Vernal Equinoctial, continuous line. 

 Autumnal " broken line. 



15. Mean Equinoctial, continuous line. 



" Solstitial, broken line. 



A. P. S. — VOL, XIV. Z 



