Meteorology of Neio Haven. 197 



E. A register kept by Prof. C. S. Lyman from Jan., 1851, to April, 

 1852, generally three times a day, and much of the time more fre- 

 quently. 



F. A register kept by Mr. HaAvley Olmstead from Dec, 1855, to 

 March, 1858, three times a day. 



G. A register kept by Mr. Joseph Bennett for a long series of 

 years. The observations are regularly made three or four, and some- 

 times five times a day. The observations taken from this record are 

 from April, 1858, to Sept., 1859, and for a part of the years 1864 and 

 1865. 



H. A register kept by Dr. D. C. Leavenworth, commencing Oct., 

 1859, and continued to April, 1864. The thermometer Avas exposed 

 on the N.E. side of his stoi'e in Chapel street, a few doors east of 

 College street, at an elevation of six feet from the ground, and about 

 forty feet above the sea. This register embraces observations three 

 times a day. 



L A few observations have been taken from a partial register kept 

 by Prof. Elias Loomis, commencing in November, 1861, and contin- 

 ued to the present time. 



K. The gaps left by the preceding journals have been filled by ex- 

 tracts from the journal of Dr. Alfred S. Monson. This journal was 

 commenced in 1821, and has been continued to the present time. For 

 the first five or six years the entries were few and irregular, but since 

 that period the entries have been made with great regularity three 

 times a day. 



The preceding registers combined extend from July, 1778, to the 

 present time, with interruptions amounting in the aggregate to fifteen 

 months, viz., from July, 1779, to Jan., 1780, and from May to Dec, 

 1795, making an aggregate of 86 years of observations to Septem- 

 ber, 1865. These observations have all been reduced as nearly as 

 possible to a uniform system. Whenever the observers adhered to 

 fixed hours of observation, the results here given are the simple aver- 

 ages of the observations with the hours annexed. When the hour of 

 observation was not invariable, an average has been taken of the times 

 of observations, as well as of the temperatures. In a few instances 

 (as in some parts of Pres. Stiles' journal) the hours were so irregular 

 that it has been thought best to reduce the observations to fixed 

 hours, by applying the corrections given in the table on page 199. 



In order to be able to reduce observations made at irregular hours, 

 to certain fixed hours, we should know the law of the hourly varia- 

 tion of temperature for each month of the year. Thiis law cannot be 

 certainly known except from hourly observations made at this place. 

 As such observations have never been made, we have endeavored to 



