DISTRIBUTION OF ATMOSPHERIC PRESSURE VON BEZOLD 367 



R. v. Miller-Hauenfels 1 , Nils Ekholm, 2 and V. Bjerknes, 3 still do not 

 appreciate the different peculiarities of these surfaces and lines as 

 perfectly as is desirable. 



Especially is this true as to researches on the processes in the upper 

 strata of the atmosphere, researches that have acquired increased 

 interest in recent times. 



One of the principal questions that comes up in the consideration 

 of an irregular distribution of atmospheric pressure is as to the 

 accelerations that the particles of air experience in consequence of 

 this irregular distribution. Most important is the acceleration in a 

 horizontal direction, that is to say, along the level surfaces of gravity. 

 Since the component perpendicular to this direction is in general 

 very small, therefore it will here be ignored. The acceleration is 

 greatest in the direction of the greatest change of pressure, that is 

 to say, in that of the greatest so-called gradient. 



This acceleration I will call the gradient acceleration and will 

 designate it by the letter j. 



In the following paragraphs we will more exactly investigate how 

 useful the two above-mentioned methods of presentation are in 

 the determination of this quantity. Especially will be considered 

 to what extent even a casual consideration of them allows of an orien- 

 tation in this direction. 



To this end it is necessary to discuss more minutely the well- 

 known formula; for this acceleration. Ordinarily one makes use of 

 the equation 



G 13.6 ,, , 



r = TTTTTT g (la) 



111 111 p 



where G is the so-called gradient, or the difference in millimeters 

 between the barometric readings that prevail at the ends of a 

 straight line 1 1 1 , 1 1 1 meters long whose direction coincides with that 

 of the greatest change of pressure; p is the density or the mass in 

 kilograms of air contained in a cubic meter; and g is the local acceler- 

 ation of gravity expressed in meters per second. The negative 

 sign that should be prefixed to this and the following formulae I 

 omit since I consider only the numerical or absolute value of y. 

 Since we cannot assume that the change of pressure is actually 



1 Theoretische Meteorologie. Vienna, 1883. 



2 Bihang till, K. Svenska Vet. Akad. Handlingar., Bd. XVI. Abt. 

 No. 5 Stockholm, iSyr. 



3 K. Svenska Vet. Akad. Handlingar, Bd. XXXI, No. 4, Stockholm, 18. 



