PHILOSOPHICAL SOCIETY OF WASHINGTON. 



109 



Table I. 



Mean amount to be added to the true difference of elevation between the summit 

 and base of Mount Washington in order to give the comptded difference, ar- 

 ranged according to the force of the wind. 



In the above table, for May, 1872, all winds under 10 and above 

 40 are included, and in May, 1873, all the cases, except a few which 

 were omitted because of serious errors in the observations. 



The table shows this remarkable peculiarity that, though with 

 winds above sixty-one miles per hour, the mean computed difference 

 in height is too great by sixty-six feet ; with winds under ten miles 

 per hour the mean difference is too small by thirty-five feet. We 

 conclude, then, that some other cause must produce the results, or 

 must act in conjunction with the wind. Taking the wind above 

 sixty-one miles per hour I have found ten cases in which the height 

 was too small by about fifteen feet, also a great number of cases in 

 which, though the wind continued stroDg from the same direction, 

 yet the computed height continually became less, showing that the 

 wind does not produce a direct effect upon the indications of the 

 barometer. On projecting the curves of pressure we find that 

 there is a uniformity in the occurrence of small and large differ- 

 ences of elevation with the maxima and minima of pressure, the 

 least being found when the pressure is high, and the greatest when 

 it is low. 



Grouping a second time, then, with respect to the maxima aud 

 minima of pressure, we have Table II. 



