444 METEOROLOGY general distribution over the United States is shown by the accompanying map (XII.), com- piled by Schott for the Smithsonian institu- tion, from all the data available in 1872. The heaviest rainfall occurs in those countries in which warm moist monsoon winds blow from the seacoast up over rapidly rising hills. Thus the annual rainfall on the Cossyah hills, facing the bay of Bengal, is 600 inches, but 20 m. further inland it is reduced to 200 inches, and at 30 m. to 100. The connection between rain- fall and the direction of the wind presents there- fore an apparent discrepancy in various parts of the world, even for stations on the same par- allel of latitude, according as the winds ascend in their course from the ocean level to hilltops, or descend from the mountains and plateaus to the lowlands. The percentage of rainy and not rainy winds is given for London in the following table, which shows that 69 per cent. of the days when S. "W. winds prevail are rainy, while only 33 per cent, of the E. and N. E. winds bring snow or rain : Relation between Rainy Days and the Direction of the Wind at London. WIND. Total number of observation!. Total number of rainy winds. Percentage of rainy winds. 8 W 1,965 1,821 69 w 982 476 51 N. W.... 1,816 754 41 N 530 191 86 N E 1134 879 83 E.. 6S9 280 83 8. E.. 722 837 47 8 876 256 68 ' Variable Calm 441 94 217
49
For the United States this relation may be pre- sented in another form. Out of 100 cases of rain or snow the average number occurring with the different winds is as follows : Percentage of Wind on Rainy Days. WIND. New Haven. Cincinnati. WIND. New Haven. Cincinnati. N... 8 2 8.W.. 15 26 N. E.. 87 10 W 1 18 E 6 1 N. W.. 7 25 8.E... 19 9 N 8 2 8 7 10 On the other hand, we have very few analyses of automatic records showing the relation be- tween the quantity of rain and the direction of the wind. The following example for Lisbon will illustrate this connection : Average annual Quantity of Rainfall for each Direction of the Wind at Lisbon. WIND. Rainfall, inches. WIND. Rainfall, inches. N... 0-61 g 8*39 N. N. E. . . N.E 0-71 0-89 8. 8.W.... 8 W 5-04 4-84 E. N. E E 0-82 . 0-43 W. 8.W.... W 5-79 2'05 E. 8. E S. E.... 1-54 I'M W. N.W.... N W 1-18 1*10 ! S 1-65 8*89 N. N. W... If 0-55 6. Barometric Pressure. It is only when we come to study the pressure of the atmos- phere, as shown by the barometer, that we arrive at a connected intelligible view of the peculiarities of weather and climate. In fact it is evident that no portion of the atmos- phere can be moved from one region to an- other, except under the influence of a pressure applied in the direction of its motion ; such movement is simply an effort to reestablish a disturbed statical equilibrium. The laws of mechanics show that relatively to the earth's surface the air would remain quiescent if the sun were absent; but the density of the at- mosphere is disturbed by the solar heat, by the variable quantity of aqueous vapor rising from the oceans and continents, and by the local condensation of this vapor into cloud and rain. In this way temporary abnormal inequalities in the distribution of barometric pressure are pro- duced, which give rise to the winds, and afford the meteorologist that connecting link which enables him to unite the whole circle of atmos- pheric phenomena into a harmonious system. (See BAROMETER.) The height of the baromet- ric column is found to vary at the same station in regular diurnal and annual changes, and also in a non-periodic manner ; on moving the in- strument to other places, its height is found to vary with the geographical position, and in an especially remarkable degree with every change of altitude. The latter class of changes have been the subject of numerous profound inves- tigations, having for their ultimate object both the utilization of the barometer for hypsomet- ric purposes, and the solution of the inverse problem, the reduction of actual barometric readings to the sea level. The solutions of both these problems can as yet be satisfactorily effected only in respect to annual means ; but within allowable limits of error approximate methods may be applied to monthly means and 6.830 0.820 atoo The dotted line is for Nertchinsk, Asia continental station ; the full line, for Tlymouth, Eng., maritime station. DIAGRAM XIII. Diurnal Variations in the Barometer. to individual observations, when the altitude of the station does not exceed 2,000 ft. The
