MONTHLY WEATHER' REVIEW. Editor, ALFRED J. HENRY Assistant Editor, BURTON M. VARNEY -- - ___- __-- -__-- ~ -~ CLOSED JUNE 3, 1925 APRIL, 1925 ISSUED JUNE, 1925 Vol. 53, No, 4 W. B. No.867 - ____ __________ - . -_ __ ___ __ - __ - __--- THE TORNADOES OF MARCH 18, 19251 By ALFRED J. HENRY The destructive tornado that swept eastward oTer parts of Missouri, Illinois, ancl Indinnn, together \vlt.li those of short.er path in Kentucky ant1 Tennessee. on March 18, 1955, created n new rec.ord of tlcst,ruct.ion 1iot.h of liuninn life rind property frtm these iiiur~h-tlrri~i.leil storms. Sewn scparatr niiil dist.inct. t.ornntlors w y e ohserved on the tlnte iiiontioned, the iiiost dest,ruct.ire of wliicli was the onc stnrt.ing. near Annapolis. Mo., which xiiovccl in an tilniost s t r q h t line bo t,he Missis- sippi R.irer, crossing tmliats st,rc:uii int.o Jnckson C'oun t.y, Ill. It laid waste u number of towns n.nd vill:i.pcs :is it, crossecl Illinois, continuing its tler-nst;iting course inbo TIIE CYCLONIC STORM TIIAT GAVE RISE TO THE TORNADOES Thn prrvious liistory of the c.3-clonic storm wibh which t,he tornnt loes mere associat.ed IS not il1umin:iting; evl- tlently thc st.orni wts i m offshoot from a cyclone whlch occupie(1 t.hc northenst Pacific from Mnrch 13 t,o 18. This c.)Pl'shoot w;w first. recognized on t h p. in. chart of t h 1Rth 11s :I depression crntererl o w r western Montmm. At t,hilt. t.imo ant1 (luring tjhc n c d 24 hours, t.liis depression gnre no cvitlence of nnj-t,Iiiiig ilut O€ thc ortlinnry ; 011 the morning of t.lw 1Sth it, W:I.S aont,crecl in northwestmm Arktmsas, ns shown in Figure 1 (A). At t,Iiis time, 7 Fit;. I.--We3t!l~r maps for 8 3. 111. :loll R p. m.. hhrch IS. 1925 3 niilcs southwest of 1,'illitEiii E. Bnrron, of t,he Cairo st'stion, and Clarence J. Root,, of the Spring- field station, were n t once cIirect,ecI to survey the ptith of blie storm. Gruteful ucknowlerlgment is here m:& for the inuttcr I hare tlraivii from the report of these twn officiicls. Iiifonantion 0.s to the remuiniw tornntloes was drnwn largely from the printed issues or" Cliinntc~logical Dntn " €or the Strites in which the storins occurred. - 1 Condmsed froin re orts tw the folloaing field omcinls: J. H. Armington. W i l l i m E. Barron James L. keudnli Roscoe Nunn. George Herder. Clmcnre J . Root. aud Geo. B. i t u r t z . with discussidn hy the editor ou the rneteoro1,oghl aspct. of the phe- nomenon. Details of I@ of life and darnpge to property, we!? included iu this REVIEW for March. 1925, and may also be found in the publlcattlon Climatologled Data" lor Mlssourl, Illinois, Indiana, Kentucky. and Tennessee for the same mouth.-Eo. 48265-25t-1 n. in. 90th nieridi:m bine, the center of lowest pressure w~ns shown ly t.he ischrs of 29.5 unil 29.i inches, respec- tivcly, lwth isol>;xrs being witAin a t.rough of low pressure that ost,entletl in IL NE.-SW. direction. A h v m t i I 1 of f 71 P c y 7 0 I I P . - Du r i ng the t l a yl igh t hours of the 1Sth thc cent-er of lowest, pressure was displnced northeastwnrcl a c1istniic.e of ahout 500 iiiiles to south- t.nstcni Inclinna, as shown in Figure 1 (B), or a t the r:\t.c of nbaut 40 milea er hour. or the purpos'c of iett,er rel:it,ing t.1ie progression. ?f the c.cnt.er of lowest pressure wit,li t,hat of the formution and progression of the turntidoes, wenther charts cover- ing t,lie lower Ohio Viilley for the hours 1, 2, 3, and 4 p. ni., central iiiericlinn time, were construcbed. The charts for 1 niid 4 1. 111. have been reprocluced in the lit.1iogriipli clinrts in kigurc 2 (A and B). 141 143 MONTHLY WEATHER REVIEW h r m , 1926 On the A chart the tracks of the tornadoes are shown by a heavy line. The times of hevinning and ending of each tornado are also shown, and &e tiiiies of beginning have been used as a basis of classification into (a) , (h) , Cc) stornis, etc. As before intimated, the pressure formation was not characterized b circular isobars but consisted of n of low pressure whose longer mis cxutended in a NE.-S’%. direction. The apparently rnpicl nioveinent of the centrnl low ressure is well understood by forecasters since the with rent rapidity toward the opposite end of the trough times carelessly considered as having a very rapid rate of translation. TAe intermediate. churts.--The c1iart.s for 1 to 4 p. ni., 90th meridian time, show rtdier conclusively that not only did the center of low pressure niove rapidly north- eastward but also that in so doing the formation RS a whole passed from that of a trough to thut of oval- shaped isobais oriented in the saiiie general direction as those of tlie earlier formation. The 1 D. ni. chart (fig. 2 (A)) show:; the lowest pressure rather restrictec 9 region of low pressure within a trou-h sout R ern center of low pressure in n trough often nioves and t 1 us the center of the formation as i~ whole is sonie- Cairo *with pressurealmost Mch. 18. 1925 as low at St. Louis and n Mch. 17. 1925 being about 200 miles northeast of Cairo and registering ressure fall, seem to indicate that the develo - great temperature contrast on t,he northern border of the niass of warm sontlierly winds at the time flowin of a westerly component aloft than at the surface. At 3 p. ni. tlie center of low pressure had assumed the form of a long narrow oval stretching from Cairo with pressure of 29.56 inches to Terre Haute, Ind., with pres- sure 29.59 inches, and tlie whole disturbance had now largely passed froni the “ trou 11’’ form to that of an oval, tlie latter being oritmted in a%E.SW. direction. The 4 p. ni. chart is reproduced as (B) in Figure 2. Lowest pressure is nom at Terre Haute, Ind., 39.55 inchm nient a greater of t P le storni was in soiiie way conditioned upon t P ie soiitheim Illinois and Indiana, which may have ha d more Over THE BAROGRAPII TRACES The barograph traces in the path of the cyclonic storni we have been considerin show not only its pro ression the atmosphere, as indicated by short irregular fluctua- tions in t.he pressure, revailed during the early niorning hours of the 1Sth. graphs were close enough to the tornado’s path to record from hour to hour, but a B so t,liat a disturbed conchion of rp one of tlie Weather Bureau baro- Mch. IS. 1925 Mch. 19. 1925 FIG. 3.-Barograph trace, Old B m cod uriue widenin out of the isobaric lines toward tlie northeast, direction. The center of the cyclone a t 1 p. in., 90th meridian t h e was robably 100 niiles or t~1iereabout.s west- 40 miles south of Reynolds County, where the tornado was first seen. If this assumption be correct and it be further assumed that the form of the inner isobar of die cyclone was that of a nort,h-soutli oval, then it ma said that the tornado probnbly developed in the nort iein left front of tlie cyclone-the northwest quadrant. The left front is a more probable place of origin tlirtn the right front since the tornado moved with greater s eed than the cyclone, and, as we shall see later, the pat P is of the two phenomena over Illinois and Indiana were nearly concurrent in point of time but not arallel in direction. The tornado moved in a direction 21 north of east while the center of tlie cyclone followed a slightly curved pat.Ii over Missouri and Illinois, concave to the north. (See Path No. VIII, chart 2, March, 1925, REVIEW.) At 2 p. m. pressure at Cairo had fallen to 29.60 inches and at Terre Haute, Ind., to 29.63 inches; whereas at Evansville, Ind., about 112 miles due northeast of Cairo, pressure had fallen to 29.65 inches only. Terre Haute thus in d icating a tendency toward movement in that soutiiwest o P Cairo, possibly in Ripley County, Mo., or be ” FIG. I.-B)nrOgrSph h c e , Cairo the characteristic oscillation due to the passage of a tor- nado. Fortunately, however, we have come into posses- sion of a baragro h trace made within less than a milt! Frankfort, Ill. This trace is reproduce? froiii tlie center o P tlie tornado that swept of Mr. J. E. Jones, city. The curve from the Cairo was about 65 i d e s due south of West reproduced (Fig. 4) as typical of the curves from other instruments near the tornado path. THE TORNADO PATHS WITH REFERENCE TO THE CENTER OF THE CYCLONE Investigations of tlie last 40-odd years have shown that tornadic stmornis rnost,l+y occur in the southeast quadrant of a general cyclonic disturbance and at a dis- tance that may range from 300 to GOO miles from the cyclone center. The editor does not recall having seen an authoritative record of the occurrence of n tornado in or very clove to tlie center of a general cyclone up to tlie presont case. It lias been suggested that the (a) tornado as de icted well toward the front of the cyclone shown on that chart. on Figure 2 (Chart -4) had its origin in the left ha P f and 143 MONTHLY WEATHER REVIEW h r m , 1926 On the A chart the tracks of the tornadoes are shown by a heavy line. The times of hevinning and ending of each tornado are also shown, and &e tiiiies of beginning have been used as a basis of classification into (a) , (h) , Cc) stornis, etc. As before intimated, the pressure formation was not characterized b circular isobars but consisted of n of low pressure whose longer mis cxutended in a NE.-S’%. direction. The apparently rnpicl nioveinent of the centrnl low ressure is well understood by forecasters since the with rent rapidity toward the opposite end of the trough times carelessly considered as having a very rapid rate of translation. TAe intermediate. churts.--The c1iart.s for 1 to 4 p. ni., 90th meridian time, show rtdier conclusively that not only did the center of low pressure niove rapidly north- eastward but also that in so doing the formation RS a whole passed from that of a trough to thut of oval- shaped isobais oriented in the saiiie general direction as those of tlie earlier formation. The 1 D. ni. chart (fig. 2 (A)) show:; the lowest pressure rather restrictec 9 region of low pressure within a trou-h sout R ern center of low pressure in n trough often nioves and t 1 us the center of the formation as i~ whole is sonie- Cairo *with pressurealmost Mch. 18. 1925 as low at St. Louis and n Mch. 17. 1925 being about 200 miles northeast of Cairo and registering ressure fall, seem to indicate that the develo - great temperature contrast on t,he northern border of the niass of warm sontlierly winds at the time flowin of a westerly component aloft than at the surface. At 3 p. ni. tlie center of low pressure had assumed the form of a long narrow oval stretching from Cairo with pressure of 29.56 inches to Terre Haute, Ind., with pres- sure 29.59 inches, and tlie whole disturbance had now largely passed froni the “ trou 11’’ form to that of an oval, tlie latter being oritmted in a%E.SW. direction. The 4 p. ni. chart is reproduced as (B) in Figure 2. Lowest pressure is nom at Terre Haute, Ind., 39.55 inchm nient a greater of t P le storni was in soiiie way conditioned upon t P ie soiitheim Illinois and Indiana, which may have ha d more Over THE BAROGRAPII TRACES The barograph traces in the path of the cyclonic storni we have been considerin show not only its pro ression the atmosphere, as indicated by short irregular fluctua- tions in t.he pressure, revailed during the early niorning hours of the 1Sth. graphs were close enough to the tornado’s path to record from hour to hour, but a B so t,liat a disturbed conchion of rp one of tlie Weather Bureau baro- Mch. IS. 1925 Mch. 19. 1925 FIG. 3.-Barograph trace, Old B m cod uriue widenin out of the isobaric lines toward tlie northeast, direction. The center of the cyclone a t 1 p. in., 90th meridian t h e was robably 100 niiles or t~1iereabout.s west- 40 miles south of Reynolds County, where the tornado was first seen. If this assumption be correct and it be further assumed that the form of the inner isobar of die cyclone was that of a nort,h-soutli oval, then it ma said that the tornado probnbly developed in the nort iein left front of tlie cyclone-the northwest quadrant. The left front is a more probable place of origin tlirtn the right front since the tornado moved with greater s eed than the cyclone, and, as we shall see later, the pat P is of the two phenomena over Illinois and Indiana were nearly concurrent in point of time but not arallel in direction. The tornado moved in a direction 21 north of east while the center of tlie cyclone followed a slightly curved pat.Ii over Missouri and Illinois, concave to the north. (See Path No. VIII, chart 2, March, 1925, REVIEW.) At 2 p. m. pressure at Cairo had fallen to 29.60 inches and at Terre Haute, Ind., to 29.63 inches; whereas at Evansville, Ind., about 112 miles due northeast of Cairo, pressure had fallen to 29.65 inches only. Terre Haute thus in d icating a tendency toward movement in that soutiiwest o P Cairo, possibly in Ripley County, Mo., or be ” FIG. I.-B)nrOgrSph h c e , Cairo the characteristic oscillation due to the passage of a tor- nado. Fortunately, however, we have come into posses- sion of a baragro h trace made within less than a milt! Frankfort, Ill. This trace is reproduce? froiii tlie center o P tlie tornado that swept of Mr. J. E. Jones, city. The curve from the Cairo was about 65 i d e s due south of West reproduced (Fig. 4) as typical of the curves from other instruments near the tornado path. THE TORNADO PATHS WITH REFERENCE TO THE CENTER OF THE CYCLONE Investigations of tlie last 40-odd years have shown that tornadic stmornis rnost,l+y occur in the southeast quadrant of a general cyclonic disturbance and at a dis- tance that may range from 300 to GOO miles from the cyclone center. The editor does not recall having seen an authoritative record of the occurrence of n tornado in or very clove to tlie center of a general cyclone up to tlie presont case. It lias been suggested that the (a) tornado as de icted well toward the front of the cyclone shown on that chart. on Figure 2 (Chart -4) had its origin in the left ha P f and APRIL, 1925 MONTHLY WEATHER REVIEW 143 The movement of the tornado across southern Illinois is definitely &sed b the time at. which the t,rdn despatchers’ The evidence of the baro- gra h traces for West Frankfort and Cniro (figs. 3 and 4), a n 8 also that of t,he time of the t.ornrtdo’s crossing into Indiana, and the Terre Haute harogra.ph, go to shorn t.li:it the two phenomena-the tornado nnd the cyclonic storni- moved very nearly concurrentJy but not parallel, the tornado in the later portion of its path being on the sout,li or the side on which waiin southerly air prevailed. Tlie remaining tornadoes developed later in t.lie rtft,ernoon and much niore distant from the cyclone cent.er, and in this they conformed to tlie experience of ninny yenrs’ stud of that phase of the phenomenon. d n t i o n may also be mncle of tlie fact tlint. t,lie forin of the isobars on March 15 mrently resembled. that on February 19, lSS4. on which Zay 44 t,ornndoes were oh- served in tbe enst Gulf States, t9he Caroliims, ancl Georgiti. On March lS, as on the d n te above nient.ioned, the t,iine of occurrence of the tornadoes was later and later in the day and farther and farther to the enstwtd AS the afternoon hours were passed. wires went out o P commission. FREE AIR METEOROLOGICAL OBSERVATIONS AND DEVELOP- MENT OB TORNADOES ON MIARCII 18, 1935 Slafione soziffi. qf thr cyclone crnfer.--Tlie records from two points are avnilable, viz, Broken Arrow, Okln., and Groesbeck, Tex. The geographical coordinates of tlie four kite stittions are given in the table below: Statim 764 4l?S i 3 Y -~ __ .. . - __ March I?’.--Southerly winds turning clockwise to W.-SW. at 3,587 rn. (11,765 ft.) prevailed at Broken Arrow; at Groesbeck the clockwise turning was also in evidence, but after reaching W.-SW. the winds bwkecl to SW. and increased in speed to 19.0 ni. p. s. at 3,750 m. (13,401 ft.)-tile t>op of the flight. The masiniuni wind speed a t Broken Arrow was 14.7 m. p. s. at 1,000 in., and but 12.8 ni. p. s. at the highest point reached. The temperatures of t,he air column above both st,a- tions were alike in that the lapse rate up to the 2 km. level was small and at both stmations there was a more or less ronounced inrersion of temperature, at ap rosi- beck. Up to these levels (3,000 to 4,000 ft.) t,he air strata were in stable equilibrium. March 18.-As tlie center of the cyclonic st.oriii passed to the eastward of the meridian of these stations the winds very nat,urally went t.o nort,herly and t.he tesi erature at beck, more distant from the center of the cyclone, also esperienced nort.herly winds, but only small changes to lower temperature, so that instead of nn inverted t.eni- perature la er in the air column there was a sniall lapse tion of stable e( uilibriuni. 17.--On this date $SE. surface winds turned clockwise to SW. at 1,780 m. (5,S40 ft.)-the to of the flight. The temperature at that level was 3.6’ 2 higher than at the surface and, as at the southern stations, there was a mate P y 1,000 m. at Broken A4rrow and 1,500 111. at &roes- Broken Arrow fell several degrees in all leve P s. Groes- rate from t c e surface to 3,000 ni. @,Si3 ft.), also a condi- Stations nort \ L o the cyclone center. Drczrl, Xarc?~ rather pronounced temperature inversion with its maxi- mum at the 1 km. level. On the following date the winds were NNW. at the surface, backing to NE. at the top of the flight, 2,765 m. (9,071 ft.j, and there was a fall in t e n erature ran mg changes are in accord with what might be especked, owing to the changed geographic posit,ion of the c.yclone center wit.li respect t.o Dresel. Royal Gknter, Mnrch IT.-This station was under the in- fluence o€ the southeastern anticyclone: accordin,gly winds were SSW. at the surface, turning to SW. and WbW. at the highest level reached, 2,845 in. (9,344 ft.). On the 18th as the cyclone center approached from the SW., the NNE. surface winds became E. at the top of the flight, 1,815 m. (.5?954 ft.1. The lapse rate on the 1‘7th was about half t.he adiabat.ic, and on the 18t,li, while it had increased sc-linewhat, W~LS yet not niore than half of the adiabatic. The chief facts brought out are t,herefore: (1) The cbsist.ence of a rather strong soutlierly current in front of the cyclone on t,lie 17th which ap arently extended the Canadian border; (2 ) the current was warm in the lower levels and cold above 3 km., with a rather small lapse rat.e between t.he surface ancl that level, therefore stable; (3) indications drawn from the barograph curves point t.0 an instability in the atmosphere in the early morning hours of the 18th and thait this inst.ahility progressed from west to east concurrently with the a~lrance of the cyclone center. Finally, the free-air records give no direct indication of t,he forces which inst.itute and maintain the tornado vort,es although the presence of a u.’arni la er of air at the 1-km. level, if liuve conspired with other conditions to pierce t e cold upper ceiling by vertical convection and thus induce a vortex which later will reach the earth’s surface. In this instance there is no warrant for assuming that such con- ditions obtained. I t must also be reluctant1 admitted that there is lit& hope that the actual con&ions that initiate a tornado vvrtcs will ever be esperinientally observed. from 5 to 15’ F. in the several leve r s. A11 of t % ese from the Gulf (.if Mexico t.0 the Great !L dies, possibly to ‘6 sufficiently warm, of w c ich me have no knowled e, may GEXERAL REMARKS The (a) tomado.-Messrs. Barron and Root, by using an autoniobile, were nble to cover the track of the (a) tor- nado in Illinois and 1ntlia.na in seven days. From its ince t,ion in Revnolcls Oounty, Mo., the tornado pur- s u e ~ P a rema~liabIy straight path through Iron, Madison, Hollinger, and Perry Counties, of that State, and across Illinois, passing t h o u h the counties of Jackson, William- course with slight deviation through the counties of Posey and Gibson in Indima, and terminated as a destructive tornado 3 miles southwest of Pet,ersburg, Pike County, Ind. The total len tli was 219 niiles with an average width Illinois, 59; and Indiana, 88. In west,ern Illinois very few observers reported the presence of a funnel-shaped c.loud; farther east., however, sonie thought they saw such a cloud, especially those who were on the outside of the storm’s path. These wit- nesses mere not very definite as to what they saw but all agreed that two clouds came together. This appearance is perhaps the most coninion testimony of esons observ- investigations on t,his type of storm were begun some 40- odd years ago. son, Franklin, H m i f ton, and White. I t continued its of less than 1 ini 7 e. 1t.s speed in Missouri was 57 m. p. h.; ing tornadic stornis. It has been repeate CY ly given since 144 MONTHLY WEATHER REVIEW APRIL, 1925 There is no doubt but that clouds are seen rushing toward a common point, viz, the vortes of t.he toriindo. As explained by Professor Dnvis inniiy years ago, it is not the rushin together of two clouds that creates the result of the whirl already in esist.ence. Root suggests that the absence of n funnel-shnped pendant cloud may be due to the fnct thnt tlie main cloud was so close to t,he earth that there was no rooiii left for the formation of the usual funnel-shnped cloud. The writer pointed out mniiy years ago thnt the chur- ncter of the pendant funnel-shaped cloud varies with geo raphic position and the average hygrometric state Thunder was heard quite .generdly a few minutes in advance of the tornado; ruin and hnil fell at va.rious Hail was moderately heavy a t Gorliani, West Pees ianhfort, nnd Carmi, Ill., and east of Pririceton irregu- lar-shaped chunks of ice as large as goose eggs were re- ported. The usual roaring sound as of severnl freight trnins ww, heard. Loss of l,$ and property.-For convenience the stntis- tics of loss of life and injury t,o persons and property loss published in the March REVIEW are here repented, with slight revision based on later estimates, in Tttble 1. TABLE 1.-Deaths, injziries, a i d properly losst-s in the stwn lortia- does of Ilfctrrli 18, 192.5 tornadic whirl % ut rather this cloud inotion is the visible of t 7 le air. . I I Root and Barron say wit.11 respect to further details of From inquiries made among the country people i t would seem that they had about five minutes’ warning after first noting the cloud. Asked as to the length of time in which thc dcstriicticin took place, opinions varied, but most persons thought aliout two minutes. If the whirl was round, tlie path of the storm 1 mile or less in width, and the velocity of translation about a milt^ a min- ute, then the tornado would pass a given point in one minute or less. There was much sameness throughout., the degree of property dsmage simply depending on what was in the track. The torLadorni tlecrenseil some- what in the lntt,er pnrt of it,s course. The (e‘) storin hegan 45 ininubcs l h r than the (c) st,orni mil 50 niilcs nlniost clue south of it, iiioving in a north- enst,crly clirect,ion. 1t.s path was from 100 to 400 yarcls wide :ind nbout. 30 miles long and it was not so severe ns t,he one immedint,ely preceding. The ig) storm lint1 ti path 100 t.0 500 yards wide and inovccl in a northettsterly clirect,ion, pnrnlleling that of t,he (e) st,orm. It was riot a st,orm of pent, intensit>y, and i t would semi tdint in the group of three storms here con- siclercd their int,ensity tliniinislied in proportion t,o thcir t1istctnc.e from the cyclone ccnt.cr. The ((7) atiti! Cf, lorn.adoas.-Tlie ((1) st,orni was first. sccn a h u t 75 miles sout,heast of t,lie point where t,he (a) t,or- nado disappeared, nncl the (f) t,ornaclo was first observed about 75 miles due soutli of where tlic (11) bornnclo dis- appeared. The path of t,he former wns 40 miles long and t,liitt of the latt,er about GO miles. The occurrence of these tornadoes, each one successively frtrtlier and farther to the sout,li and closely relnt,ed in point of time, suggests an aun!ogy between the origin of scconclary cyclones and tornncloes. It is a mntt,er of comiiioii knowledge that when a cy- clonic syst,em entering the continentm from the Pttcific cttn not progress enstwitrcl dong the. nort.liern boundary by reason of t.he presence of unt,ownrd atmospheric condi- tions x secondary cyclone will almost always develop to the southward of the primary. Similarly, it, is conceiv- able tlint the in) tornndo of this series yacluJly €ound itself in at.niospheric surroundings whir 1 macle its fur- ther enclumnce impossible : hence tdic whirl lost contact with t.lw enrt,li ant1 soon disappeared. Fn.rther south, where t.he atmospheric conilitims evident,ly were more fit\-oruble, n second whirl developed, mntle contact wit,h the eart,li, ant1 it, too, disuppenrecl after tt course o.f 40 miles. In ench case movement toward thc northenst, broumlit the whirl into regions of lower t,eniperature and, possihy, less moisture cont,ent. This action wns agnin rcpcated in the case of the (f) tornado, whicli originabed in Marion County, Icy., and moved AS beforc describcd. The ohservntinns on the (cl) tornado were especially worth wliilc ; blicy confirm the suggestion by Root, pre- viously mentioned, with respect t,o t,lie lower end of the pendant funnel cloud. Following is a brief account. of this torn:do, with escerpt,s from ICenclall’s report. t,reme western Inc P innn and about 3OCJ miles dist,nnb. .~PRIL. 1935 MONTHLY WEATHER. REVIEW 145 The tornado originntecl in Harrison, County, Ind., nioved thence enst,-nort>lie.zst. hi a ath about 10 miles on t.he pendant funnel cloud is discussed as follows: * * * The c!istriot. t.ravcrsed rises in a roIIing'pIateaii, with t,hc highest part a rat.lier ahrupt escarpiiient on t.he eastern edge, along the Ohio River. whcre t,he elevation averages al.>out. X50 feet above sea level. With the iiicreasing elevation the funnel of the t.ornado becanie more and more deeply truncated, which caused the paOh of prscticxlly tot.al destruction to widen to about, half a mile. At the edge of the platenu along t,he Ohio River, where the descent is very abrupt, being about. 400 feet in as many yards, the funnel (lipped down ininiediately and destroyed all builc1ing.s i n it,s path, even at the foot of t,he blufi over which i t hac1 come. At t,his point the path narrowed t.o aljeut. 9 0 feet.; it. narrowed fiirt.lier in crossing the Ohio, but. widened again t.0 more than 1,000 feet as higher ground was reached ahout 3 miles to the east.wsrd, after which i t generally cont.ract.ecl until i t wLw only 60 feet wide at long and of vaiying width. The c B! ect of rising groiintl Pewee Valley. * * * At. the time of its passage the center of the low-presslire area was near Indianauolis. Inimediatelv after the uass::ge of the tornado a t Louisvilie the skies cleared; tlie air b e c h e calm, and the teniperatiire rose about 8' * * *. Mr. Kendall notes that t81iis tornado passed through Harrison County near Elizabeth, wit.liin less t,hm 2 miles of the pn.tJi of the severe tornado of May 3 i 1 1S90, which, it map he remembered, s t,ruck Louisville, causing greiit loss of life and roperty. The Louisvil r e barograph shows frequent oseillthons froni about, i :3 0 a. 111. to 4 p. M.. or just, hefore tho paw sage of the tornado. Tlie lowest point on t.he L o u i 4 l e trace was reached a t G p. ni., wh~ch corresponds pretty closely with the time of the p:wsage of t.he cyclone cent.cr about 115 niiles to the nort.11. dBILITY OF MODERN STRUCTITRES TO WITHSTAND TORNA1)OES Much inberest, is evident in recent y c a ~ in tlie il.hilit,y of mell-const.i~icteil builrlings of brick, st,one, concret,e, or what not, to mithstancl the t.crrific force of the wind as esertecl in tornadoes. t.he building was swept from it,s fountlntion. The damage to school buildings of brick construction articularly noticeable; the roofs were ripped off and WLs t 7 le upper stories badly wrecked. A correspondent of Engineering News-Record, writing in the issue of March 86, menbions the fact that a t Mur- physboro, while A clean sweep was made of structures in the northwest part of the city, two reinforced concrete coal bins within 300 feet of the Mobile & Ohio Railroad slio s, which latter were destroyed, were left standing uniniaged in t.he niidst of n mass of wreckage. Also near t,he railroad shops two steel wheat bins * * * are still intact, although one of them is leanin The was claniaged considerably. Immediately in the rear is a 1 60-foot reinforced-concrete smokestack. I n spite of the great amount of destruction around this structure, it remains standing and shows no signs of damage what- ever. A small two-story building of plain concrete was practically destroyed, its 8-inch walls being sheared off entirely a t t,he top of the fiivt floor. Root ant1 Barron, in a supplpment,al report on damage t,o buildings, say: F r a m tliirelli,rgs.--ZTnless well built, largely totally demolished in inain path of tornado. A house in Griffin, Ind., lying on its side was returned t.IJ its original position by workmen practically intact. It had diagonal sheathing, which added much strength. Of houses not destroyed, tlie roofs and porches were taken off and in some cases the second story. R f i m o reuiderrces.-An architect in Murphysboro invited our at.teiition to the fact that stueco houses resisted the storm to best advantage, and we found froin observat.ion that they did stand up I)et.t.er t h i i franie hildings. There were few sducco houses esrept in Miirphysboro. Brick biiilrli,i0.s--rf;ciioolu.-For the most part in two-story brick svhools the first floor walls rcniained practically intact; in the sec- ond story the iiitcrior wills largely remained standing, though the oiit,er walls criimblecl. The Mobile C Ohio shops at Murphys- horo, brick I)iiilclings, were dcmolislicd by wind and afterwards hirncrl. Iir general, brick store buildings in the direct path of the st.orm wcrc dcstroved. A new brirk two-story mine office building at Orient No. 3 t h e at West Frankfort was practically undam- R ~P C I . but i t was in the lee of the large steel niine tipple. To the bent of our meinory, brick buildings shod rip where they had steel t,russerl roofs. Steel co)LXt.rirr!in,i.-~t.eel water and oil tanks belonging to the r:iilrn:id at. Gorhani were unharnied. A rsiinilar steel water tank at. West Frankfort niiiir was blown over. At, the same niine (Orielit No. '3) the steel conveyor was badly damaged, but the large iiiodern st.eel tipple was not greatly harmed. The tipple at Caldwell mine (wood and steel) was demolished. brick building of the Brown Shoe Co. in Murp E- ysboro ( We saw none.) THE TORNADO OF APRIL 5, 1925, NEAR MIAMI, FLA. By RICHARD W. GRAY [Wcnthrr Burmu Oficc, n1i:mi. Fln.. April 15, ICE51 The destructive torna.ilo which p.zsse(1 north of Miami during t,he early afternoon of Sundsy, April 5, 191'5, occurred in connection wit,li n rlisturhnnce t,hitt. h ~t d nioretl soutlienst,wnrd RC~ORS t.he ITnit,etl Stat,es from t'lie California coast nntl that was central over cstrciiic northern Florida a t the time of t>lie t.ornado. Tlie t,ornaclo develo ed over t.he Evergl,ztlcs, apparen tly in the vicinity of Hi;% P enh. which is about, 4 miles north- west of t,he city liiiiits of Minnii nntl nhout, S miles norlli- west of the Weather Biirenu station. Tlie funnel cloud wa,s first, o'lxwrvd by goli pliiyers on the municipal golf coii~sc at Hislenli at 1 p. 111. or n few minutes earlier. It8 tlevelopincnt, wa.s rtlso swn by niaiiy ot,lier pcrsons whose nttention !i d been a ttmcted by the unusunlly threatening sky which att,entletl n. thundeirst,orni ilnd lidstmin preceding the t.orn:do. The opport.unit.ies for observing the storm were esreptionnlly favorable. The usual large Sunday crowcl W:I.S out of cloois, nisny hundred nut.oniobi1ist.s lwing near the t,ornndo path. Moreover, on nocount of its slow progress, word of the tornado was w-itlely s read and several thousand pel-ons watched i t until it, cisappearet~. Mnng ohservers st,ntecl bliat the development of the t,ornndo iiiimediately followed the unit.ing of two dense cloud nin.ssos. When first. seen by t,he writer, a t 1:15 p. ni., t,lic development WIW complete, and tlie funnel cloud nppenred as i i very slender cone est,encling in a st.might line from blie dense cloud niass above to the earth. With t,lw cxuoept,ion of a slight bending and t,wist.ing of the lower mrt of the cone, there was no funnel cloud. This was undouhtedly due t.o t,he slow movement of the general cloud niass. Tlie funnel cloud, howt~er, frequcn tly rose from the ground only to descend again within 11 few niinubes. When its end t,ouched the ground t,hrre inrnriably followed a phenomenon siniilar tlcvintion at) any h i e 1 roiii the vertical position of the