224 



Meteorology of Nexo Haven. 



The following table shows the mean temperatui-e of each month of 

 the year as deduced from all the preceding observations. Column 

 9th shows the mean temperature of each month corrected by the 

 table on page 199, after having received the modifications hereafter 

 explained. 



We next endeavored to determine hoAV these results might be em- 

 ployed to correct the assumed table of the mean diurnal change of 

 temperatui-e given on page 199. The numbers of that table for each 

 month were set oif as ordinates corresponding to the hours as ab- 

 scissas, and a curve as nearly symmetrical as possible, was drawn 

 through the extremities of these ordinates. Then taking the results 

 for January, given above, the morning temperature was subtracted 

 from the mean of the month, and this difference was set off as an 

 ordinate upon the January curve, by which means Ave determined 

 one point of the true diurnal curve for that month. In like manner, 

 the mid-day and the evening observations determined two other 

 jDoints of this curve. In the same manner, three points were deter- 

 mined in the diurnal curve for each of the other months of the year. 

 The points thus determined corresponded so Avell with the assumed 

 table on j^age 199, that we were encouraged to make a further attempt 

 to determine the true diurnal curve for New Haven. We therefore 

 commenced anew a reduction of all the journals which embraced 

 observations made at not less than five hours of each day. These 

 observations embraced a period of nine years, and the hours of ob- 

 servation were generally about 6 and 10 a. m., 2, 6, and 9 p. m. The 

 following tables show the results of these observations. 



