July 25, 1890.] 



SCIENCE. 



(44) May 12, 1886. Attica, Fountain County, Ind. (scale 



3). — Time, 10 p.m.; iu vicinity, 9 killed; 200 houses 

 razed; loss, 1200,000. 



(45) April 15, 1887. St. Clairsville and Martin's Ferry, 



Belmont County, O. (scale 3). Time, 3.20 p.m. ; none 

 killed; about 200 buildings of all kinds demolished; 

 loss, $250,000. 



(46) April 21, 1887. Prescott, Linn County, Kan. (scale 3). 



—Time, 5.30 P.M.; 20 killed, 237 wounded; 330 build- 

 ings razed in vicinity ; loss, $150,000. 



(47) April 22, 1887. Mount Carmel (near), Wabash Coun- 



ty, 111. (scale 3 — ). — Time, 6 P.M.; 2 killed, several 

 wounded; every thing in path destroyed; loss, $50,- 

 000. 



(48) April 22, 1887. Clarksville (near), Johnson County, 



Ark. (scale 3).— Time, 6.30 A.M. ; 20 killed, 75 to 100 

 injured in vicinity; loss, $150,000. 



(49 ) June 16, 1887. Grand Forks, Grand Forks County, 



Dak. (scale 3). — Time, 3.22 P.M. ; 4 killed; 50 or more 

 houses, besides hundreds of barns, etc., razed; loss, 

 $150,000. 



(50) Feb. 19, 1888. Mount Vernon, Jefferson County, 111. 



(scale 3-(-).— 18 killed, 54 wounded; 100 buildings 

 razed; loss, $400,000. 



(51) May 27, 1888. Hillsboro, Hill County, Tex. (scale 



3—).— Many buildings razed; loss, $100,000. 



(52) Aug. 21, 1888. Wilmington, New Castle County, Del. 



(scale 3).— 1 killed, 20 wounded; loss $100,000 to 

 $200,000. 



(53) Jan. 9, 1889. Brooklyn, Kings County, N.Y. (scale 



3).— Time, 7.40 p.m. (Eastern); width, 500-600 feet; 

 length, 2 miles; whirl from right to left; roar heard 

 10 or 15 minutes before; loss. $300,000. 



(54) Jan. 9, 1889. Reading, Berks County, Penn. (scale 3). 



— Time, 5.40 p.m. ; swept from west to east in a path 

 60 to 100 feet wide; wind often seemed to crush 

 from above; 40 killed; loss, $200,000 (estimated). 



(55) Jan. 12, 1890. St. Louis, St. Louis County, Mo. (scale 



3). — Time, 4 p.m. ; moved to northeast in a path 500 

 to 2,000 feet wide; heavy rain for 3 minutes; greatest 

 damage where path was narrowest; 3 killed; 100 

 houses razed; loss, $250,600. 



(56) March 27, 1890, Metropolis, Massac County, 111. (scale 



3— ).— 1 killed, 50 injured; loss, $150,000. 



(57) March 27, 1890. LouisviUe. Jefferson County, Ky. 



(scale 3-|-). — Time, 7.57 p.m. ; path at beginning 600 

 feet, as it left the city 1,500 feet; cloud did not quite 

 reach the earth; great damage to property; 76 killed, 

 200 injured; loss, $2,250,000. 



This list comprises all the most destructive storms that 

 have been reported, as far as a definite locality was men- 

 tioned. It has been found exceedingly diflBcult to deter- 

 mine the loss in many cases, because an estimate has evi- 

 dently been made of the loss to crops, orchards, etc., from 

 the rain, hail, and floods that accompanied the tornado, and 

 not from the wind itself. Again, the loss reported evidently 

 referred to a large region in the county, and not to any 



specific town, 

 lows: — 



45 



Some of these may be enumerated as fol- 



It is highly probable that in some of these cases the losses 

 from one county have been estimated in another, though 

 this has been avoided as much as possible. It is very much 

 to be hoped that more definite estimates will be made in the 

 future. The loss to structures by the wind should be care- 

 fully distinguished from the loss of every other kind, by 

 hail or flood, and to crops, stock, or orchards. I shall be 

 very grateful to any who will send me corrections to this 

 list, or add other tornadoes. H. A. Hazen. 



LIGHTNING-CONDUCTORS FROM A MODERN POINT OF 

 VIEW.' 



A LIGHTNING-CONDUCTOR used to be regarded as a conduit or 

 pipe for conveying electricity from a cloud to the grouud. The 

 idea was, that a certain quantity of electricity had to get to the 

 ground somehow; that if an easy channel were opened for it the 

 journey could be taken quietly and safely, but that if obstruction 

 were opposed to it violence and damage would result. This being 

 the notion of what was required, a stout copper rod, a wide- 

 branching and deep-reaching system of roots to disperse the 

 charge as fast as the rod conveyed it down, and a supplement of 

 sharp points at a good elevation to tempt the discharge into tliis 

 attractive thoroughfare, were the natural guaranties of complete 

 security for every thing overshadowed by it. Carrying out the 

 rain-water-pipe analogue, it was natural also to urge that all 

 masses of metal about the building should be connected to the 

 conductor, so as to be electrically drained to earth by it; and it 

 was also natural to Insist on very carefully executed joints, and 

 on a system of testing resistance of conductor and " earth," so as 

 to keep it as low as possible. If ever the resistance rose to 100 

 ohms, it was to be considered dangerous. 



The problem thus seemed an easy one, needing nothing but 

 good workmanship and common sense to make accidents impos- 

 sible. Accordingly, when, in spite of all precautions, accidents 

 still occurred; when it was found that from the best constructed 

 conductors flashes were apt to spit oS in a senseless manner to 

 gun-barrels and bell-ropes, and wire fences and water-butts, — it 

 was the custom to more or less ridicule and condemn either the 

 proprietor of the conductor, or its erector, or both, and to hint 

 that if only something different had been done, — say, for instance, 

 if glass insulators had not been used, or if the rod had not been 

 stapled too tightly into the wall or if the rope had not been made 

 of stranded wires, or if copper had been used instead of iron, or if 

 the finials had been more sharply pointed, or if the earth plate 

 had been more deeply buried, or if the rainfall had not been so 

 small, or if the testing of the conductor for resistance had been 

 more recent, or if the wall to which the rod was fixed had been 

 kept wet, etc., — then the damage would not have happened. 

 Every one of these excuses has been appealed to as an explanation 



' By Professor Oliver J. Lodge (from Industries). 



