6o 



SCIENCE. 



[Vol. XVI. No. 391 



that balls of fire were playing about it. The highest wind 

 recorded at the Signal Office, less than 1,800 feet from the 

 path, was 36 miles per hour. After it had passed, the wind 

 shifted suddenly to the west, and continued to blow from 

 that point for twenty-four hours, and with increasing velo- 

 city, the record showing 42 miles per hour at 9 p.m. The 

 sky was perfectly clear at 9.30 p.m., except a streak of very 

 high and apparently motionless cirrus in the west. Shortly 

 after this time the atmosphere became obscured by a peculiar 

 haze or smoke, through which the moon shone with a red- 

 dish light. 



"Where the tornado entered the city, the width of the 

 path was a little more than 600 feet. As the cloud progressed, 

 the width of its path increased to 1,500. There is no evi- 

 dence that the tornado-cloud touched the ground at any 

 point in Louisville. In nearly every case the destruction 

 was confined to the upper fioors of the demolished buildings, 

 but comparatively few houses being totally ruined; and also 

 a large proportion of the one-story structures in its path were 

 uninjured. Churches, halts, warehouses, and other struc- 

 tures having but little interior support, sufi'ered the most. 

 To this fact is attributed the principal loss of life. At the 

 Falls City Hall alone, where a large number had congre- 

 gated, 44 were killed. Frame buildings invariably with- 

 stood the shock much better than those constructed of ma- 

 sonry. But few of the destroyed buildings bear evidence of 

 being actually blown down by the whirl of the tornado- 

 cloud itself, but their destruction was apparently caused 

 rather by a lateral or vertical rush of air-currents centring 

 toward it. The ruins and the disposition of the debris give 

 ample evidence of this. The right [south] side of the storm- 

 track, and in a less marked degree the left side, afforded 

 numerous examples of the intensity of this lateral force. In 

 both cases [i.e., on either side of the track] the sides of the 

 buildings facing the storm were pulled out, the debris falling 

 towards it [the track]. In many cases fragile articles, such 

 as glassware, remained undistui-bed. In the centre of the 

 track the destruction was mainly due to a vertical force 

 which lifted the roofs of the buildings. The roof of the 

 Union Depot was lifted bodily, and deposited intact on the 

 floor, immediately beneath its original location. 



"The destroyed buildings were, as a rule, of a very un- 

 substantial character, being mainly ordinary brick dwell- 

 ings, small stores, and warehouses. The Fort Nelson 

 building, at 7th and Main Streets, is the most notable excep- 

 tion to the general destruction in the tornado path. This is 

 a well-constructed six-story building, and, by its greater 

 height than those surrounding it, was more exposed to the 

 storm's fury. Despite the fact that it was directly in the 

 storm's track, and that all other houses on either side were 

 wrecked, it escaped with the loss of its windows. The 

 gyratory motion of the tornado is well illustrated in the dis- 

 position of the prostrated trees in the parks, and the timber 

 outside the city. In the centre the trees were piled in pro- 

 miscuous heaps, denoting a tremendous wrenching or twist- 

 ing force. On the right side the tree-tops point almost 

 north-east; those on the left side, nearer due east. Through- 

 out the path of the storm the zone of destruction on the 

 right [south] side is more than twice as wide as that on the 

 left side, and shows a much greater intensity of force." 



I have quoted almost the whole of Mr. Burke's most ex- 



cellent report, which may be regarded as a model in its 

 terseness and statement of valuable facts. One or two com- 

 ments are added. The heavy rain preceding the tornado 

 seems to have been independent of it, and not necessarily 

 due to any action in it. The presence of lightning, and 

 absence of thunder, are noteworthy, and may explain the 

 fact that oftentimes no electric action is reported in a tor- 

 nado. Thunder is always regarded as the most prominent 

 characteristic in a thunder-storm. The sudden shifting of 

 the wind to a direction from the tornado is very significant. 

 How could this have been if there were a partial vacuum in 

 the stoi-m ? The failure of the gaslights is very interesting, 

 and at first sight appears to show a marked diminution in 

 pressure. Diligent inquiry at a gas-office, however, shows 

 that no reasonable diminution in the pressure of the atmos- 

 phere could affect a gaslight; moreover, this same phenom- 

 enon was noted at Cleveland, O , at a time when there was 

 a very high wind, but no tornado. Investigation has shown 

 that a high wind forces the immense gas-holder against the 

 vertical posts, and this causes the failure in the pressure. 

 This explanation is now accepted by those familiar with the 

 subject. It seems also probable that such a high wind, 

 when blowing near the earth, would have a tendency to an 

 upward thrust; and this has been shown repeatedly by the 

 uplift of dust, leaves, and small branches in front of a sud- 

 den gust. The increasing wind after the tornado was not 

 connected in any way with that, but was simply the high 

 wind in the rear of a storm, and in this case it was intensi- 

 fied by the steep barometric gradient. The preservation of 

 the Fort Nelson building was probably due to a diminution 

 in, or a lifting-up of, the storm just at that point, and not 

 to its construction. There are innumerable instances in 

 which the frailest structures have been left undisturbed in 

 the centre of a tornado-track. The evidence from the 

 prostrated trees does not bear out the gyratory theory, but 

 is rather strongly against it. If there were any gyration, 

 say, from right to left, the trees on the left (north) side 

 ought to face west, and not east as they were reported. 



Mr. Burke, in a pi'ivate correspondence, states that he 

 made a most diligent search for evidence of corks being 

 blown from bottles, but did not find a single case, although 

 there were several apothecary-shops and bottling-establish- 

 ments in the centre of the track. He also states that the 

 diminution in the gaslights was quite sudden, and passed 

 away quickly. It would be interesting if some one would 

 determine the velocity of the wind needed to produce this 

 effect. This report must be regarded as corroborative, and 

 not as absolutely establishing points now in doubt. The 

 statements in the report show only very slight evidence of 

 bias or leaning toward any theory; and this is just what is 

 needed now, if we are to make any headway in our studies. 

 The report is worthy of commendation. 



Incidentally one of the more interesting facts brought out 

 by the tornadoes on this date are barograph traces made by 

 the passage of the tornado at Owensboro, Ky., and Cincin- 

 nati, O. These are given in the accompanying plate. Figs. 

 6 and 7. Unfortunately the original trace for Fig. 6 was 

 carried by the owner to Scotland, and given to Mr. Buchan, 

 who rightly regarded it as a most valuable acquisition. It 

 is to be regretted that the original was not kept in this coun- 

 try, where it belongs, and a copy taken abroad. The copy 



