The sad story of malicious tampering on America's railroads from ICC accident reports

While preparing the page on accidents caused by throwing a switch irrationally in front of an oncoming train, I included for contrast a couple of accidents where the switch was maliciously misaligned. This made me wonder about how many accidents have been due to malicious actions, so I searched the ICC accident reports for the adjective "malicious," and discovered 95 reports. Some of these included the phrase "no malicious intent," and so were not germane. 78 reports remained, however, in which malicious intent was proved or suspected, dating from 1914 to 1958. These provided a depressing picture of such events, which are reviewed here for what lessons they contain.

The word "sabotage" is not used often in ICC investigations (only two cases were found), but it does describe precisely these events. The word sabot means "wooden shoe" in French, and sabotage was originally just the word for the manufacturing of wooden shoes. However (so I understand) the wooden keys used to hold rails in chairs in the English fashion, which were widely used in France, are also called sabots. If you have a hammer, you can knock the keys out fairly easily and displace the rails. In the United States, the equivalent action is pulling spikes with a claw bar. This was done by saboteurs, and now the word, especially in English, refers to malicious tampering and destruction, usually as a clandestine act of war, or figuratively. I'll use the word here for brevity.

Of the 78 cases of probable sabotage, more than half, or 45, involved tampering with switches (oddly, the word "turnout" is rare in ICC reports). In 34 of these, the switch was "open" or aligned for the divergence. An unsuspecting train might derail at the divergence, if it were moving rapidly (above 40 mph for the usual No 10 turnout), or more probably collide with equipment on the diverging track. In 11 of them, the switch was only "cocked," which means that the points were moved an inch or so from the stock rail so that the flange would drop between them and derail immediately. The next most common form of sabotage, in 11 instances, was tampering with the track, usually to disassemble a joint and displace one of the rails, or to unspike a rail and move it sideways. In 9 cases, objects were put on the tracks. This is seems to be a specialty of juveniles, and in most cases the objects are swept aside or noticed and removed without causing an accident. However, a craftier version puts objects on the high rail of a curve, where the flange pressure is ready to cause derailment. There was one accident due to an engine running away, sent on its way by malicious intent, and one case of a landslide at a cut that may have been caused by a dynamite explosion. One of the track accidents mentioned above, on an electric line near Niagara Falls, also involved dynamite to damage the track.

There are similar accidents that do not involve malicious intent. One is the careless leaving of switches open or unlocked by trainmen, rather common accidents usually affecting yard and switching movements. Another is poor maintenance of switches, of which there was one example only in the ICC reports (CStPM&O, Mendota, MN 14 Mar 1914). There were two examples of switches run through incorrectly in the trailing direction. If this is not detected at the time, the switch is rendered hazardous to trains approaching in the facing direction, since the points are usually free to move or distorted. These cannot be detected on the train running through the switch unless one looks. In one of the two cases, the train derailed was instructed to look out for a switch run through that the block signals had detected, but a sufficient lookout was not maintained (PM, Glen Lord, MI, 28 May 1929).

In most cases, an open switch can be detected from its switch light or target, if the weather is clear, in time to prevent an accident. In automatic block signal territory, a circuit controller operated by a rod from the switch points, sets the signals to stop when the points are not in the proper position. A break in a rail also interrupts a track circuit and protects the fault by the block signals. These protections can be defeated by a crafty wrecker, and sometimes by not maintaining good switch lights where there are no automatic block signals. The usual American manual block system also offers no protection against open switches or broken rails, since there are no track circuits or other electrical safety features.

The most determined wreckers smash the switch lamp lenses, turn down the wick into the reservoir, and bend the target so it shows clear. The switch lock is sawed off, the points moved slightly and wedged with a stone or something. The rod to the circuit controller is detached by removing a cotter pin and a nut, so that it does not give protection. Although these measures are usually characteristic of an adult rogue, a 6-year-old once found a switch unlocked and amused himself by opening it, but could not close the switch properly again BR&P, Salamanca, NY, 21 Feb 1927). Three boys smashed the lock of a switch on the Piedmont and Northern (P&N, Greer, NC 22 March 1955) and replaced it. Later, a person unknown lined the switch for a coal dock, diverting a freight train with the loss of two lives.

Main line switches are locked by stout padlocks that require some effort to remove, unless you happen to have a switch key. Switch locks have been removed from the vicinity to create the impression of malicious tampering when the fault lies with negligent trainmen (C&O, Harvey, WV, 6 Nov 1929).

Tampering with the track is almost always the work of adult men, since it requires considerable physical power to pull the spikes, unscrew track bolts, and slew rails. These rogues realize what the bond wires at the joints do, and carefully do not break them. The current use of continuous welded rail prevents this sort of sabotage almost perfectly, except at what few rail joints remain. Outstanding examples of this are the accidents on the SP at Harney, NV on 12 Aug 1939, and on the PRR at Baden, PA, 16 Mar 1941. 9 passengers were killed in the first accident, and 2 in the second. The perpetrators were unknown, but the evidence points to disgruntled employees.

Juveniles of limited mental capacity enjoy putting junk on rail heads, and this is aided by the practice of section crews to leave odd material lying around on the right-of-way. Certain areas became known for this (NYC&StL, Cleveland, OH, 19 August 1932), and they should be carefully cleaned of junk so that there is no enticement. As we have mentioned above, adult perpetrators know how to use such junk to good effect, by placing it on the high rail at curves. On the Southern at Brennan, GA a 10-year-old put a rock on a track that derailed a train on 14 Feb 1933. A similar event occurred at Milltown, WI on 24 Feb 1948, where a Soo passenger train was derailed by stones on the high rail at a curve, that derailed the engine truck of the locomotive. Three 11-year-olds were responsible. At Bloomfield, IN a 10-year old put a bolt and nut on the rail that derailed an Illinois Central freight train, killing 3 employees, on 25 August 1951. At Walton, IN on 27 January 1947 two boys dragged a roll of fence wire, some pipes and other hardware, all found nearby on the right of way, onto the track. Engine 5377 of No. 207, with 8 cars, running at 60 mph, wedged the junk into a turnout and derailed. Three passengers and an employee were killed.

Many cases of obstructions on the track are not sabotage, but have other causes, such as a blown-down tree, or bulk cement carelessly spilled. A search of the ICC reports with the search phrase "obstruction near track" yields a surprising 290 documents. Of course, a number of these say no obstruction on the track, or words to that effect.

Vincent Williamson was a notorious young train-wrecker of 1927-1928. At the age of 17 he escaped from the home for the feeble-minded at Salem, OR on 1 Jun 1927 with his buddy Herman Lemp and took to riding trains. Hobos were not created by the depression--they existed previously, and were a dangerous, violent mob of barbarians, not worthy of being made folk heroes. By August, Williamson and his buddies made it to Topeka, where they shot and killed a Union Pacific special agent who had ordered them off the property. They were put off a Chicago and Alton freight in February of the next year, and wrecked a passenger train at Independence, MO on the 15th of February by putting objects on the high rail at a curve. On the 24th, Williamson and a "negro" companion wrecked a Katy train at Parsons, KS by tampering with the track. Williamson wanted revenge for being put off at Fort Scott, the "negro" for being put in jail (I use the original words in the ICC report; it would seem unjust to say African-American). They claimed they wanted to rob, but did not do so. On 7 March, Williamson and an associate Thaddeus Atkins put objects on the Union Pacific tracks at Lenape, KS, but failed to cause an accident. They did, however, succeed in being arrested. On 9 March, they earned 5 to 10 at the Hutchinson Reformatory, and on 26 March at Erie, KS they added 10 to life. One hopes that, in the event, it was closer to life. Note how little Williamson's victims were valued by their society.

Malicious mischief was not confined to the track and switches. At Gallitzin, PA, the summit of the Pennsylvania Railroad's crossing of the Alleghenies, there was sabotage in December, 1931 that involved closing the angle cocks on the brake lines of trains about to descend the steep grades in either direction. This area was the site of derailments due to excessive speed in 1921, 1925, 1927 and twice in 1947. In one of the 1947 derailments, the 13th and last car of a passenger train, the 10-5 sleeper Cascade Mirage, became detached due to a defective coupler. Unfortunately, the hand brake was also defective and the brake reservoir was depleted, so the car derailed on a 9° 15' curve 3.37 miles east of Gallitzin. The other accidents, all to passenger trains, occurred on the notorious Bennington Curve, an 8° 30' curve 1 mile east of block station SF at the eastern portal of the Gallitzin tunnel.

In the present case, Extra 4272 had reached the summit with 88 cars of coal, 7200 tons, with 3537 and 4463 as pushers. Between MO and AR, west of the tunnel, the conductor and brakemen turned up retainers for the descent. These would retain a certain pressure in the brake cylinders, keeping the brakes applied, while the engineman released the brakes and recharged the reservoirs. Speed was limited to 8 mph between MO and AR, and was supposed to be 12 mph through the tunnel to SF, but the speed was generally a bit higher. The descent begins before the west portal, so the brakes were applied in the tunnel, on the 1.39% descending grade. The engineman found very little exhaust, and discovered that the brakes were applied only on the engine, tender and about 12 cars of the train, hardly enough to be noticed on the 2.36% descending grade past SF. He whistled for brakes, but was not heard.

A brakeman and the conductor were still on top of the train at this time, making their ways back to the cabin car. The flagman on the caboose did not open the conductor's valve, and by the time the conductor got there, it was too late. The train derailed on Bennington Curve, a mile east of SF, at 7.09 pm, moving at over 40 mph.

It was discovered that the angle cock at the rear of the 12th car was closed. This was done some time after the brake test at Gallitzin, where the brake pipe pressure was raised to 100 psi, and the minimum of 85 psi was noted at the cabin car. The brake pipe behind the 12th car was properly charged, and if the conductor's valve had been opened promptly, there was probably sufficient brake power to stop the train. It was surmised that leakage had caused the brakes to creep on anyway, retarding the train more than the few brakes at the head end could do.

Earlier, a westbound train had experienced trouble stopping at Gallitzin, and angle cocks were found closed. There had been several other incidents in the days before the event, none of which had had serious consequences. Two boys were seen hanging about, and they were probably responsible for the mischief.

Direct sabotage of signals seems rare. Most malicious tampering with signals, such as cutting wires, would result in the display of the most restrictive aspect. It would require more intelligence than generally possessed by the average villain to cause a false clear. Of course, where there are automatic block signals, wreckers are careful not to disturb bond wires and to disconnect circuit controllers so the signals will not warn of the danger.

The stimulus for nearly all of the cases of sabotage was either individual revenge or evil thrills and pleasure. Most of the perpetrators are unknown. Of the 78 accidents, only 12 were "solved." There were no claims of credit afterwards, so they were not intended as public demonstrations. Robbery was not a motive (except on the T&P at Mackle, TX 24 Nov 1922, where a "negro" extra section laborer cocked a switch so he could loot the derailed freight train). Dismissed employees sometimes turned rogue, as at Osgood, NC, 19 Jan 1915, on the SAL. Many of the track and switch tampering cases exhibit knowledge of railroad devices and operations that are not common knowledge. It is evident that the motive in most is sheer malice.

The accident tolls fell most heavily on engine crews, who were innocent victims of the wreckers. Very few passengers were killed in all of these accidents taken together, which is the only comfort that can be derived. These accidents have continued, and several more recent accidents could be added, such as the sabotage west of Phoenix a few years ago on the SP.

For the period 1911-1960, Shaw mentions two of the accidents included here, Henryetta and Harney. For the times before the ICC reports, he mentions 5 accidents, the first in 1868. Of course, the sabotage that occurred during the Civil War is properly excluded. If the same ratio of 2 to 78 is applicable to the earlier period, we can estimate a total of 273 incidents of malicious tampering from 1830 to 1966, about two per year on the average. Shaw concentrates on accidents that involve passenger fatalities, while most of the fatalities in these accidents are of the unfortunate engine crew, as we have already pointed out. This probably explains the disparity between Shaw's few mentions, and the results of my computer search.

In contrast, Rolt's book on British accidents does not mention a single case of sabotage or malicious tampering in the same period. Access to the right-of-way may be slightly more difficult than in the United States, and there are more eyes and less isolation, but graffiti show that this is no barrier, especially in the urban areas where one might expect malicious behavior to be more common. There is also the almost complete absence of the tramp element, which has no doubt been responsible for a significant number of American incidents.


The ICC accident reports can be found at DOT Website. The year 1960 is absent.

R. B. Shaw, Down Brakes (London: P. R. MacMillan, 1961).

L. T. C. Rolt, Red for Danger, 3rd ed. (Newton Abbot: David and Charles, 1976).

Return to Railway Index

Composed by J. B. Calvert
Created 22 February 2002
Last revised 5 March 2002