464 Observations respecting a new Scale 



deduced by adding the number in e to the parallel num- 

 ber in C. 



F is the greatest heat in tJw sun of each latitude A, de- 

 duced by adding the number in f to each of the parallel 

 puinbers in E., 



d \s an increasing series of ninety, in number corre- 

 sponding \yith the degrees of latitude from to 90, in co- 

 lumn A, viz. from 5 to 95. 



£• is a decreasing series oi nine, in number from latitude 

 to 90, viz. from 30 to 21. 



y is a decrtasi?ig scries of 7iine, in number from latitude 

 P to 90, viz. from 35 to 26. 



g is a decreasing series of nine in number from the midr 

 -^lle latitude of 45 up to 90, and the same down from lati- 

 tude 45 to 0— latitude 45, being 13. 



The temperatures marked in columns D, E, F, give the 

 temperatures of ordinary years ; in order to give the defec^ 

 or Gxceas of extraordinary seasons the parallel number in 

 g must be subtracted, or added to either of the columns 

 rnD,E,F. 



Mr. Kirwan has given the mean annual temperature of 

 every latitude, from the equator to the north pole : in this 

 table each tenth only is given, but the temperature of each 

 intervening latitude may be obtained easily by calculation 

 from those given, viz, by allowing one- tenth for each der 

 gre& of latitude, in the columns c?, e,f, p^, 



N. B. The thermometer, in taking observations, is sup- 

 posed to be placed at about 5 or 6 feet from the ground, 

 and perfectly insulated or detached from any body which 

 can cause reflected heat, when pdaced in the sun j and for 

 observations in t^e shade, placed ip a north aspect, where 

 the sun never reaches. 



Hencp, according to this calculation, the greatest cold at 

 the poles, or the greatest natural cold, is at —68 of Fah- 

 renheit, or 100 degrees below the freezing point of water ; 

 therefore, in order to reduce Fahrenheit's scale to this, 68 

 inust be added to every degree of Fahrenheii-s above the O5 

 and for any degree helotp the p of Fahrenheit, that nun^- 

 ber must be subtracted from 68. 



It might be expected, as the gradations in the above 

 table are unifofm, and calculated from the equator to the 

 poles, that the temperatures in thjs table should correspond 

 with considerable exactness with actual observ^tior>s in 

 pvery instance; but it should be recollected, that a dif- 

 ference must occasionally arisp from various causes, viz, 

 ' ' . ' difTerence 



i 



