198 



NATURE 



[July i8, 1918 



accompanying figure (Fig. 3) is a freehand composite 

 illustration, showing in a broadly generalised way a 

 weather map characteristic of tornado occurrence in 

 the Central Mississippi valley region of the United 

 States. Tornadoes also spring up under conditions 

 which differ considerably from those here illustrated. 

 It is, therefore, impossible to select or to draw any 

 fixed "tornado-type" map. 



Protection of Life. — ^The possible protection and pre- 

 servation of human life in tornadoes are very real 

 and vital questions over large areas of the United 

 States. From a long and intimate study of tornadoes 

 Finley deduced certain rules for the protection of life 

 which have over and over again proved their accuracy 

 and value. If a tornado is approaching, from west or 

 south-west, and the observer is on or very near its 

 probable path, the best thing to do, if there is time, is 

 to run north.. "Dug-outs" or tornado-cellars should 



-Composite weather map, showing conditions favourable for tornadoes K. DeC. Ward), 

 From the Quarterly Journal of the Royal Meteorological Society. 



be provided near the house. The safety secured bv 

 means of "dug-outs" is that rhey remove' persons who 

 seek refuge in them from risk of injury from flying 

 debris, also from the danger of being picked up by 

 the winds. 



If there is no time to escape, or if escape is im- 

 possible, the safest place is to stand, face forward, 

 against the west or south wall of the cellar, as near 

 the south-west corner as possible. The reason for 

 these precautions is this : that the debris of the house 

 will, if the building is destroyed, be most likely to be 

 carried towards the north-east. Hence north-east or 

 east rooms and walls are least safe. If caught out- 

 doors, and otherwise unable to escape, the best thing 

 to do, as a last resort, is to lie flat on the ground in 

 an open space, face downwards, the head to the east, 

 and the arms placed over the head for protection. 



Protection of Property : Tornado Insurance. — In 

 regard to the protection of property certain things are 



NO. 2542, VOL. lOl] 



fairly clear. Tornadoes cannot possibly be prevented ; 

 and no building, certainly none of any practical use, 

 can be built to withstand the violence of the wind in 

 the vortex of a well-developed tornado. Hence the 

 only resource left is to protect life and property to 

 the best of our ability and with a knowledge of the 

 facts which have been brought to light by a sane, 

 unprejudiced, scientific study of the phenomena. 

 Owing to the varying intensity of tornado violence 

 and of the velocity of the surface winds, the damage 

 done to different sorts of buildings varies greatlv. If 

 the intensity of the storm is not sufficiently gceat to 

 destroy everything in its path, the damage done bv 

 the less violent winds will obviously depend largely 

 upon the strength of construction and upon the build- 

 ing materials. It was Finley's advice to build "as 

 you would without the knowledge of a tornado." He 

 found, however, that, other things being equal, a 

 frame building seems to resist 

 destruction better than one of brick 

 or stone. The modern steel-con- 

 struction buildings have some of 

 the "elastic" quality which renders 

 frame structures safer than the 

 more stable and solid ones of stone 

 or brick of the older style. It 

 makes little or no difference in the 

 end whether a building is in a valley 

 or on a hill. 



In view of the property loss occa- 

 sioned by tornadoes it is natural 

 that tornado insurance has become 

 a widespread and popular method 

 of financial protection. So far, 

 however, the business has not been 

 carried on upon ' a thoroughly 

 scientific basis. Tornado insurance 

 to the amount of several hundred 

 millions of dollars is carried, largely 

 by general fire insurance companies 

 and partly by local mutual insur- 

 ance companies. The definition of 

 a tornado is usually crude and un- 

 scientific, and there is much un- 

 necessary confusion. It is true that 

 the more conservative companies do 

 prohibit some " risks," such as 

 windmills, old and frail buildings, 

 large plate-glass windows,' and the 

 like. It is' interesting to note the 

 marked rise and fall of the amount 

 of tornado insurance with the oc- 

 currence in any year of severe or 

 destructive tornadoes. Closely fol- 

 lowing the St. Louis tornado of 

 May, 1896, there was an increase of tornado 

 insurance of nearly 10,000,000 dollars, and after 

 the Omaha (Nebraska) tornado of Easter Sunday, 

 1913, several million dollars' worth of tornado 

 insurance was written in Omaha and the sur- 

 rounding districts, which were at once thoroughly 

 canvassed by insurance agents. Many new "dug- 

 outs " and cellar caves were built at the same time. 

 As Prof. H. E. Simpson * has pointed out, tornado 

 insurance risks differ from others in several ways, 

 notably in the fact that there is no criminal hazard 

 present. For people cannot remove, or explode, or 

 destroy their buildings for the sake of the insurance 

 on the plea that the damage was done by a tornado. 

 It is obviously wise to scatter tornado risks across, 

 not along, the usual path followed by tornadoes. 



* H. E. Simpson, "Tornado Insurance." Monthly Weather Review, vol. 

 x.\xiii.. pp. 534-39. (Washington, D.C., December, 1905.) (A short biblio- 

 graphy i< appended.) 



