July 26, 1888] 



NATURE 



301 



when the train was close to the seismograph, and con- 

 tinued visible until the train had passed off the bridge at 

 the other end. 



DOES PRECIPITATION INFLUENCE THE 



MOVEMENT OF CYCLONES? 

 T N Prof. Elias Loomis's first " Contribution to Meteoro- 

 1 logy," in the American Journal of Arts and Science, 

 he examined the distribution of rain around 152 storms 

 (cyclones) in the United States, in order to determine 

 whether there exists any relation between the velocity of 

 a storm's progress and the extent of the accompanying 

 rain area. He found that " the average extent of the 

 rain area on the east side of the storm's centre is 500 

 miles ; and when the rain area extends more than 500 

 miles, the storm advances with a velocity greater than the 

 mean ; but when the extent of the rain area is less than 

 500 miles, the storm advances with a velocity less than the 

 mean." In his twelfth "Contribution" he examined 39 

 storms which moved with exceptional velocity (1000 miles 

 or more per day) and found that " the rain area generally 

 extended a great distance in advance of the storm centre, 

 the average distance being 667 miles." Finally, Loomis 

 examined 29 cases of those abnormal cyclones in the 

 United States which moved toward the west. He says : 

 — " In nearly every case we find a fall of rain or snow in 

 the region toward which the low centre advanced, and in 

 most of the cases the rainfall was unusually great. . . . 

 It may be inferred from these comparisons that the fall 

 of rain or snow is one of the most important causes which 

 determine the abnormal movements of areas of low 

 pressure " (ninth memoir, p. 44). Ley and Abercromby 

 state that in Great Britain the relation of the weather to 

 the cyclone centre is the same whatever the path of the 

 cyclone ; thus when storms advance toward the west the 

 greatest cloud development and rainfall is to the west of 

 the cyclone centre. In the Proceedings of the Royal 

 Meteorological Society, vol. xliii., Abercromby gives a table 

 showing the relation between the intensity of " trough 

 phenomena " and the velocity of cyclones. This table 

 indicates very clearly that the greater the velocity of the 

 cyclone the more marked the "trough phenomena." 

 Hence, according to Abercromby's definition of " trough 

 phenomena " the heaviest rain and cloud areas are massed 

 toward the front of rapidly advancing cyclones, while 

 immediately after the passage of the line of minimum 

 pressure the sky begins to show signs of clearing. This 

 is especially well marked in cyclones passing off the north- 

 east coast of the United States. When the cyclones are 

 moving with unusual rapidity, not only all the rain, but 

 almost all of the cloud area is confined to the front half 

 of the cyclone. 



Loomis suggested that the excess of rain in front of 

 rapidly advancing cyclones was one of the causes of the 

 rapid advance ; but when investigating heavy rainfalls in 

 the United States he concludes that " the forces which im- 

 part that movement to the air which is requisite to an 

 abundant precipitation of vapour, instead of deriving in- 

 creased strength from the great volume of rain, rapidly 

 expend themselves and become exhausted;" and after 

 examining certain cyclones which were accompanied by 

 no rain he adds : " So that it seems safe to conclude that 

 rainfall is not essential to the formation of areas of low 

 barometer, and is not the principal cause of their forma- 

 tion or of their progressive movement." Hann arrives at 

 similar conclusions from investigations in Europe. After 

 investigating an especially heavy rainfall which occurred 

 in Austria and vicinity in August 1880, he concludes 

 thus : — " The appearance of a barometric minimum in 

 Hungary occasioned abnormal and extended precipitation 

 on the west and north-west side of this barometric de- 

 pression. The reaction of this precipitation on the position 



of the centre of the depression is scarcely perceptible. . . . 

 We find, therefore, through the investigation of the relative 

 lowest barometer reading in its behaviour to rainfall, that 

 our former conclusions are confirmed " (lxxxii. Bunde d. 

 Wiss. ii. Ab., November 1880). This investigation does 

 not necessarily prove that precipitation does not appre- 

 ciably influence the movements of cyclones in general, but 

 at least suggests that in the first cases mentioned above 

 the unequal distribution of rain around rapidly moving 

 cyclones was not the cause, but the result of the cyclone's 

 advance. In cyclones which move very slowly, as do 

 tropical cyclones, the air ascends almost uniformly around 

 the centre ; but when cyclones have a more rapid pro- 

 gressive motion, the air in the rear, which has not only to 

 enter, but to follow the cyclone, is more retarded by 

 friction than the air in front, and hence does not enter 

 the cyclone so freely, so that the formation of cloud and 

 rain in the rear is retarded ; while, on the other hand, a 

 larger volume of new air enters the progressing cyclone 

 in front, and increases the amount of precipitation. Thus, 

 between February 12 and 14, a cyclone passed across the 

 American continent with the exceptionally high velocity 

 of 58 miles per hour. During its passage the highest 

 wind velocity reported on any of the United States 

 Signal Service morning weather maps was 40 miles per 

 hour, occurring immediately in the rear of the cyclone at 

 Father Point, Can., on the morning of the 14th. At none 

 of the other 130 stations did the maps show a wind velocity 

 exceeding 30 miles per hour during the passage of the 

 cyclone. This is an example of many similar cases which 

 show that in rapidly moving cyclones the air in the rear 

 near the earth's surface does not move as rapidly as the 

 cyclone itself. Hence, it seems evident that the air near 

 the surface immediately in the rear of these cyclones is 

 not air which has followed the cyclone near the surface, 

 but air which has descended from above. Espy showed 

 many years ago that, on account of mechanical heating by 

 compression, no descending air can be accompanied by 

 precipitation ; and an explanation is thus afforded why 

 there is none, or but little cloud and precipitation in the 

 rear of rapidly moving cyclones. On the other hand, in 

 order that a cyclone may advance rapidly, there must be 

 a rapid decrease in pressure, and consequently a rapid 

 removal of the air, in front of the advancing depression. 

 Since, according to the normal circulation of a cyclone, 

 there is an inward movement near the earth's surface 

 and an upward and outward movement near the top, this 

 upward and outward movement is necessarily increased 

 in unusually rapid-moving cyclones, and hence also the 

 cloudiness and precipitation are increased. 



Hourly observations of cloud movements made during 

 the day hours for nearly two years at Blue Hill Observa- 

 tory indicate that the velocity of storm movement, and 

 especially the variability of the weather, are intimately 

 connected with the velocity of movement of the general 

 atmosphere. 



The writer is hence led to believe that the main cause 

 of rapid cyclone progression is an unusually rapid drifting 

 of the atmosphere over large regions ; and the unequal 

 distribution of rain around the cyclone is due to the 

 rapid progress of the cyclone. 



H. Helm Clayton. 



Blue Hill Observatory, Boston, June 18. 



NOTES. 



Mr. John Whitehead returned to Labuan in safety from 

 his second expedition to Kina Balu, and is daily expected in 

 England. He ascended the mountain to its summit, and at- 

 tained to an altitude of 13,500 feet. His collection will contain 

 many novelties, the small portion sent by him in advance to 

 Mr. Bowdler Sharpe exhibiting many curious features. The 

 new species will be described by Mr. Sharpe in the forthcoming 



