228 Mr. J. P. Harrison Lunar Influence on Temperature. [May 4, 



cloud and vapour is due to the sun and the winds, and not in any wise, 

 as it would appear, to the moon, or, lastly, to that system of compensation 

 and alternation which seems to ohtain so frequently in atmospheric pheno- 

 mena, and is so suggestive of mechanical force. 



The exceptions to the rule of a higher temperature occurring at the 

 moon's first quarter, and lower temperatures after full moon, in any single 

 year or group of lunations, are not more frequent than occur during the 

 annual march of the seasons, and affect the position of the mean hottest 

 and coldest day in the solar year. 



Several curves besides those referred to in the text are appended. 



Description of the Curves. Plate IV. fig. 1. Curve of mean tempera- 

 ture for 618 lunations (1814-65), from the Greenwich observations as 

 corrected by Mr. Glaisher. 



Fig. 2. Curve of mean temperature for 99 lunations (1856-64), from the 

 same source. 



Fig. 3. Curve of minimum temperature from the Greenwich observations 

 for the same 99 lunations. 



Fig. 4. Curve of mean temperature for three years, or 37 lunations 

 (1859-61), at Oxford, from the photographic curves of temperature taken 

 at the Radcliffe Observatory. 



Fig. 5. Curve of mean temperature for the same three years as in fig. 4, 

 from the ordinary means of the days at Greenwich, to compare with fig. 4. 



Plate V. fig. 1 . Curve * of mean temperature for 520 lunations 

 (1814-56) at Greenwich. 



Fig. 2. Curve of mean temperature for the 86 lunations (1841-47) 

 during which bihourly observations were taken at Greenwich. 



Fig. 3. Curve of mean temperature for 86 lunations (1837-43), from the 

 Ordnance observations at Dublin. 



Fig. 4. Curve of mean temperature at Oust Sisolsk (Siberia), for 86 

 lunations (1837-43), to compare with fig. 3. (Mean of Russian observa- 

 tions at 18 h , 2 h , and 10 h .) 



Fig. 5. Curves of minimum temperature for one year (1859) at Green- 

 wich and Utrecht. 



Note. In 1848-56, the maximum occurred on the second day after first 

 quarter, and a second maximum before last quarter. The minimum was 

 found on the third day before first quarter, and the second minimum on 

 the day before full moon. 



* This curve appeared in the British Association Reports for 1859. 



