METEOROLOGY. 541 



is due the earlier occurrence of the niorniug maximum aucl the un- 

 usual depression of the afternoon minimum. At night time the aii 

 flows back again, whence results the diminution of the morning mini- 

 mum in the valley, but its increase on the flank and summit. An 

 analogous process occurs between land and sea; during the day the 

 air at an altitude above the land flows towards the sea and causes there 

 a rise of pressure that shows itself even on the coast by the delay iu 

 the morning maximum and the afternoon minimum. The reverse pro- 

 cess occurs at night. This is illustrated in the California barometer 

 curves and those deduced by F. Chambers for the seven English self- 

 recording stations. The elucidation of a general theory of the diurnal 

 barometric oscillation is much facilitated by having clear views of the 

 modifications that this oscillation experiences by reason of the periodi- 

 cal diurnal transfer of air from the land to the sea, and from the valley 

 to the hills. The daily barometric oscillations attain an extraordinary 

 extent in the excessive summer heat and dryness of Arizona ranging 

 from 1.8 mm. above to 2.2 mm. below the daily mean, or an oscillation 

 of 4.0 mm. {Z. 0. G. M., xvn, p. 35.) 



Koppen, in a study upon the vertical (and horizontal) distribution of 

 the pressure in the atmosphere, discusses the allowable simplification 

 of Euhlmann's formula for approximate reductions of observations to sea 

 level or other altitudes. He states that since August, 1880, when his 

 formulae, viz — 



B_ h h 



b ~ 18460 + 72 (t + jL (45°-<p)) ~ 72 (256 + t + T \ (66Q.5- cp)) 



were presented to the meteorological committee at Berne, he has used it 

 iu the form of a manuscript table for all cases occurring at the Deutsche 

 iSeewarte, where greater accuracy was required, although for the daily 

 weather reports the reductions are as before, made without reference to 

 the prevailing air temperature. 



He then proceeds to discuss the important question of isobars and 

 gradients for higher levels as revealed by reducing barometric readings 

 upward to an assumed level (say 5,000 or 10,000 feet), the need of which 

 has been felt these many years before by the present writer and of late 

 by others. We desire, namely, to know at what elevation within an 

 area of low pressure the gradient ceases to be inward and above winch, 

 therefore, the air must be flowing outward, or at what altitude' the 

 pressure above a low area becomes equal to that of the same altitude 

 above a neighboring high area, or again at what altitude and distance 

 from the center of a depression the pressure above any station in the 

 quadrant of cold northerly winds becomes equal to the pressure above 

 another station in the quadrant of warm southerly winds. 



Let the lower temperatures and pressures be t and f 1? B and B, ; the 



