Mean Direction of tlie Whid. 245 



Avas employed, the record showed a permanent change, indicatirxr 

 either a change in the elasticity of the spring, or a change in the ap- 

 paratus by Avhich the motion of the pressure ])late was transmitted to 

 the recording pencil. 



The observations on the force of the wind connnenced 1857, Se])t. 

 7d, 7h, and continued to 1859, July 11th. Until the month of Apiil, 

 1858, pressures less than ten ounces seem to have been recordetl 

 with as great fidelity as the higher pressures. Alter A[)ril 5th, 1858, 

 no pressures were recorded less than ten ounces on a plate ten inches 

 square. About this time, either the spring or the recording apparatus 

 must have sustained some injury. It is impossible now to determini' 

 why the apparatus subsequently failed to lecord the smaller pressures ; 

 nor can we determine whether the higlier pressures recorded before 

 April, 1858, are comparable with those subsequently recorded. This 

 failure of the anemometer to record the low pressures impaii's some- 

 what the value of the observations ; nevertheless, the results are so 

 consistent with each other, and Avith similar observations made else- 

 where (as we shall see hereaftei;), that the observations are considered 

 worthy of preservation. 



Other observers have experienced similar difficulties with the press- 

 iire apparatus of Osier's anemometer. At the Observatory of Toronto, 

 Canada, during the years 1840, '41 and '42, in pressures of less than 

 one pound, the pressure plate of the anemometer either did not move 

 at all, or the record of its motion was very imcertain. In higher 

 winds the instrument worked well, but the spring was insufficient to 

 bring the pencil back again to the zero, so that luitil corrected by 

 hand, the pencil might continue to mark high pressures after the wind 

 had lulled. A similar impei-fection was found in the Osier's anemom- 

 eter employed at the Girard College Observatory in 1840-45. 



Table VIII exhibits in detail the entire series of observations at 

 Wallingford, and shoAvs the recoi'ded force of the wind estimated in 

 ounces upon a surface of 100 square inches, for each hour of the day 

 during tAvo years. The average force of the wind is thence obtained 

 for each hour of each montli. 



17 



