August I, 1889] 



NATURE 



Z2>^ 



level. A third condition favourable to an abundant rainfall is 

 proximity to the ocean, especially when the prevailing wind 

 comes from the ocean. Capes and headlands projecting con- 

 siderably into the ocean generally show a rainfall greater than 

 interior stations only a few miles distant ; and lastly, Prof. 

 Looniis notes that the great and non-periodic depressions of 

 the barometer are always accompanied by a considerable fall of 

 rain, and that the average tracks of these depressions are marked 

 by an excess of rainfall. 



The following are some of the conditions unfavourable to 

 rainfall. Fresh winds blowing in a nearly uniform direction 

 throughout the year, such as prevail within a portion of the 

 system of trade winds, especially in mid ocean. The rainfall 

 on Ascension Island is quoted as a case in point, observations for 

 two years showing that the direction of the wind was south- 

 east or very nearly so during the time, the rainfall during these 

 two years being 2 '31 and 4 '30 inches respectively. This con- 

 dition of things prevails over the Atlantic Ocean within the 

 region where the trade winds blow with considerable force and 

 are seldom interrupted. 



A second condition unfavourable to rain is a position on the 

 leeward side of a range of mountains running in a direction 

 nearly at right angles to that of the prevalent wind. An illustra- 

 tion of this principle is seen on the Malabar coast of Hindustan. 

 On the ocean side of the range of mountains the rainfall is 250 

 inches annually, whilst on the eastern side of the range the air 

 is very dry, and the amount of the mean annual rainfall is less 

 than 25 inches. 



When there is a second range of mountains, parallel and 

 within 100 or 200 miles of the first, the influence of this cause is 

 considerably intensified, and this diminution is still more decided 

 when a place is surrounded by mountains, or nearly so. Sala- 

 manca is so situated, and the mean annual rainfall there is less 

 than 10 inches. 



Elevated plateaus have generally less rainfall than insulated 

 mountain peaks of an equal elevation ; this is illustrated by the 

 fact that Leh, being situated on that remarkable plateau of 

 Tibet, has a ;r.ean annual rainfall of less than 3 inches. Another 

 similar case is found in the tableland (the Punos) between two 

 great chains of the Andes ; and it is observed that the average 

 height of the Sahara being over 1500 feet, this elevation may 

 contribute in some degree to the smailness of the rainfall. 



Another condition unfavourable to rainfall is the dryness of 

 the atmosphere, under which head Piof Loomis gives three 

 special cases, viz. remoteness from the ocean, measured in the 

 direction from which the prevalent wind proceeds, areas of 

 high barometric pressure, and high latitudes. This last conclusion 

 does not state that the average rainfall regularly diminishes as 

 we go northwards, the same as the mean temperature ; but if the 

 mean annual fall be t^ken for every 10° of latitude the important 

 influence on the amount of rainfall is very decided, and is 

 emphatically exhibited in high latitudes. The general table of 

 observations, arranged in order of latitude, which began this 

 chapter, shows that for the four stations whose latitude exceeded 

 71° the mean annual rainfall was 7*44 inches, whilst the 

 paucity of observations of the fall of rain or snow that have 

 been made during the various Arctic expeditions also demonstrates 

 the fact. 



A review is next made of the regions which show a very small 

 rainfall, and the causes inquired into where the observations of 

 pressure, temperature, and humidity of the air, and the direction 

 and force of prevailing winds rendered it possible to obtain 

 some definite information as to the meteorological condition of 

 the region. 



Prof. Loomis has thoroughly investigated the conditions of 

 rainfall in the United States, and from the tables of observations 

 he arrives at the inference that the depression of the barometer 

 accompanying extraordinary rainfalls is not very great, the 

 average pressure at the low centre being 29 '63 inches for the 

 part of the United States north of latitude 36°, and the average 

 pressure at the stations of greatest fall being 2977 inches. 



Table cxii. gives all the cases where' the barometer fell below 

 29 inches at any station in the United States or Canada between 

 September 1872 and June 1884, and also the station where the 

 greatest rainfall occurred for the preceding twenty-four hours. 

 The conclusion drawn from such a comparison is that a moderate 

 depression of the barometer is as favourable to great rainfall as 

 an extremely great depression, which would seem to indicate that 

 rainfall has but little connection with barometric depression. It 

 must, however, be remembered that the depression at the centre 



of a low area depends not merely upon the barometric gradient^ 

 but upon the geographical extent of the low area. 



The following are some of the conclusions Prof Loomis arrives 

 at respecting the causes of rainfall in the United States. One of 

 the most common causes of rain is an unstable condition of the 

 atmosphere resulting from an unusually high temperature com- 

 bined with unusual humidity. Another very common cause of rain 

 frequently associated with this is a cold northerly or westerly 

 wind in the western segment of the low area, and proximity to 

 the ocean or to a large inland sea. 



The investigation affords important evidence respecting the 

 influence of rainfall upon areas of low pressure, viz. — 



No great barometric depression with steep gradients ever 

 occurs without considerable rainfall. 



In great rain-storms the barometric pressure generally 

 diminishes while the rainfall increases. 



The greatest depression of the barometer generally occurs 

 about twelve hours after the greatest rainfall. 



A great fall of rain is favourable to a rapid progress of the 

 centre of least pressure, while a small rainfall is generally 

 attended by a less rapid progress. 



It is also noted that some of the characteristics of areas of 

 low pressure with little or no rain are : — 



(i) Feeble barometric gradients. 



(2) Moderate winds. 



(3) Slow changes of barometric pressure. 



(4) Slow progressive movement. 



Whilst in similar areas of low pressure with excessive rainfalls 

 all these conditions are reversed. 



In order to study the influence of rainfall upon barometric 

 pressure under different geographical influences, Prof. Loomis 

 has compiled for Europe a similar set of tables to those concern- 

 ing the United States. Of the 106 stations having a rainfall of 

 not less than 2 "5 inches in twenty-four hours, eighty-six are 

 situated south of latitude 48°, and fifteen are north of latitude 

 48°, indicating that heavy rains are about six times as frequent 

 in the south as in the north of that latitude. Prof. Loomis 

 thinks that the summary of observations relating to Europe 

 seems to indicate that great rains occur on the west side of the 

 low centre more frequently than they do in the United States. 



Tables have also been prepared showing the rainfall over the 

 North Atlantic as far as observations permitted. An unexpected 

 fact exhibited by these tables is the prevalence of rainfalls with 

 the barometer somewhat above 30 inches. 



A comparison of the results that have been obtained for the 

 United States and for Europe brings Prof Loomis to some im- 

 portant conclusions respecting the influence of local causes in 

 modifying the relation of rainfall to barometric pressure. 



The conclusions are, for stations east of the Rocky Moun- 

 tains : — 



(i) South of latitude 36°, a rainfall of 2 J inches in eight hours 

 at any station occurs on the east side of a low area more 

 frequently than on the west side in the ratio of 2 '6 to i. 



(2) North of latitude 36°, a rainfall of 2 inches in eight hours 

 at any station occurs on the east side of a low area more 

 frequently than on the west side in the ratio of 28 to i. 



(3) A total rainfall of 9 inches in eight hours at all the stations 

 east of the Rocky Mountains occurs on the east side cf a low 

 area more frequently than on the west side in the ratio of 6 '2 

 to I. 



(4) Over the North Atlantic Ocean great rain areas occur on 

 the east side of an area of low pressure more frequently than oi> 

 the west side in the ratio of 2'6 to i. 



(5) In Europe a rainfall of 2\ inches in twenty-four hours at 

 any station occurs on the east side of a low area more frequently 

 than on the west side in the ratio of 2'0 to I. 



These results indicate that in the United States and Europe,, 

 as well as over the North Atlantic Ocean, great rainfalls are 

 generally associated wi'h a barometric pressure somewhat below 

 the mean, and the precipitation occurs chiefly on the easteriv 

 side of a low area. 



The relation of a rising to a falling barometer with rain points 

 to the conclusions that at Philadelphia the amount of rain which 

 falls while the barometer is descending is nearly three times as- 

 great as that which falls while the barometer is rising. The 

 entire Atlantic coast of the United States north of latitude 36°' 

 exhibits results similar to those found for Philadelphia. Ad- 

 vancing westward from the Atlantic coast, the ratio of the ])re- 

 cipitation when the barometer is falling, compared with that 

 when the barometer is rising, changes somewhat rapidly, and 



