May 2, 1890.] 



SCIENCE. 



271 



first thing seen at a long distance, is not seen till it is just 

 upon the observer. This is due to an enormous cloud of 

 dust directly in front of the tornado, rising oftentimes to a 

 great height. Such a cloud of dust has o''ten been seen in 

 thunder-storms in regions never visited by the tornado. It 

 is produced by a tremendous outrush of air from the tor- 

 nado, and indicates almost conclusively that there is a 

 plenum, and not a vacuum, at the centre. That this out- 

 rush of air could be produced by the fall of raindrops carry- 

 ing air with them, has heen proved impossible. The calcu- 

 lation showed that the heaviest rainfall that ever occurred 

 could not produce a velocity greater than one-tenth of a 

 mile per hour, which is practically inappreciable (see Ameri- 

 can Meteorological Journal, September, 1887, pp. 206-211). 



Loud Roar. 



As the tornado approaches, an indescribable roar is heard. 

 It has been likened to the bellowing of a million mad bulls, 

 the roar of ten thousand trains of cars, etc. This is certainly 

 a most significant fact, and one that has not been sufficiently 

 dwelt upon. The roar was analyzed by one observer, and 

 was found to be precisely similar to a continuous roar or 

 rumble of thunder. There is no question but ihat it is a 

 marked electrical phenomenon, though just the manner of 

 its production demands most careful investigation. Another 

 explanation will be given under another heading. 



The Tornado. 

 These warning sights and sounds are quickly followed by 

 the funnel-cloud itself, like a great balloon sweeping its 

 neck round and round with terrible fury, and destroying 

 every thing in its path. It has been likened to an enor- 

 mous elephant's trunk. It whirls with almost incredible 

 velocity in its mad career, with a motion back and forth, 

 sometimes leaving the earth a moment, then bounding back 

 to continue its dire havoc. The whole destruction occupies 

 but three or four minutes; but in that time the stanohest 

 houses of brick or stone have been demolished, and sorrow 

 and ruin have been spread all along its path. 



Clearing Sky. 



The exceedingly circumscribed nature of the tornado is 

 shown by the blue sky or stars appearing a few minutes after 

 it has passed The wind turas to the south again, every 

 thing quiets down, and no one would think that a terrible 

 catastrophe had occurred, except for the devastation that is 

 •all about, and the cries of the unfortunates. 

 Width of Path. 



The width of the destructive path has been given as high 

 -as a mile, but this is undoubtedly due to an erroneous esti- 

 mation of the real track. There are always south-westerly 

 indraughts which produce more or less destruction on the 

 south side, but these should be caref uUj distinguished from the 

 .track proper or the region of greatest destruction. This may 

 reach a thousand feet, though its width is rarely over two 

 hundred or three hundred feet. 



Distribution of Debris. 



In the central line of the tornado all timbers and trees are 

 tstrewed in the same direction, as though a mighty river had 

 passed and left them behind. Where trees are not com- 

 pletely uprooted on either side of the path, they all lie with 



their tops inclined to the central line; that is, on the south 

 side they are toppled toward the north-east, while on the 

 north side toward the south-east. Tais appearance or action 

 may be due to the indraughts, and should be carefully dis- 

 tinguished from the effect of the tornado proper. The same 

 remark may be made as to the distribution of fences and 

 light objects on either side of the path. 



Velocity of the Destructive Wind. 

 This should not be confused with the onward motion of 

 the tornado, which is comparatively slow. The whirling of 

 the cloud is the cause of the great destruction. This has 

 been estimated as high as a thousand miles per hour; but 

 such a velocity as that is highly problematical, and is due to 

 erroneous assumptions. Probably the mostaccurate measure- ' 

 ment of this velocity ever made was at Wallingford, Conn., 

 on Aug. 9, 1878. Here the tornado blew off monuments in 

 a cemetery without chipping either the upper or lower stone. 

 In one case the stone was 2X2X4 feet in dimensions, and 

 would have required a velocity of about two hundred and 

 sixty miles per hour to blow it off. 



Direction of Whirl. 



This has been reported again and again as from right to 

 left, "counter-clockwise;" and this has caused a notable 

 modification of the old theory that the whirl in a tornado 

 may be either way. It is probable that this unanimity of 

 opinion is due to other causes than careful observations. 

 The determination of the direction of a whirl is practically 

 impossible, unless the observer is within a very short dis- 

 tance, as any one may see for himself in the whirls of a 

 dusty street. The only possible way to determine the direc- 

 tion of the whirl is to be near enough to see the actual mo- 

 tion of the leaves and twigs upon the ground. If one looks 

 at the whirling column, he can learn nothing positive. This 

 question is by no means settled yet, and there are reasons 

 for thinking the old theory correct. 



Air-Pressure. 



Until very recently it has been assumed that there is a 

 partial vacuum at the tornado centre. It is admitted that in 

 a thunder-storm the pressure rather suddealy rises, nearly a 

 tenth of an inch oftentimes; and, if the tornado is analo- 

 gous, we may reason that the pressure rises in that. Pro- 

 fessor Davis, in the American Meteorological Journal (Feb- 

 ruary, 1890, p. 452), says that this fails to take account of 

 the causes of increased pressure in thunder-storms, and of 

 the decreased pressure in tornadoes. Tliis seems a remark- 

 able argument. Where does the severest thunder-storm and 

 greatest increase of pressure leave off, and the tornado, with 

 an absolutely reversed pressure, begin? The strongest argu- 

 ment that has ever been advanced has been the seeming 

 bursting or exploding of houses in a tornado. This could 

 have been occasioned by a sudden blow on any side of the 

 house, whereby the pressure inside would be increased, and 

 the walls thrown outward. This view was advanced some 

 years ago, and has received a recent most remarkable con- 

 firmation in the report of a destructive storm at St. Louis, 

 Jan. 12, 1890. The observer, a very intelligent man, writes, 

 "In all cases of tailing walls, it was noted that they fell 

 outwards. In all cases of houses which at first sight ap- 

 peared to have exploded, it was ascertained that immediately" 



