The Wharton Safety Switch

The motivation for the design of this switch and how it worked.

Railway vehicles are guided by their track, so some arrangement to permit a vehicle to choose one or another route is a fundamental requirement. In the US, the device is called a turnout, from the old term used on tramways. There are two parts to a turnout. One part allows the flanges of a wheel to pass through the other rail. This is the crossing, or frog, the latter name an American term arising from the appearance of the device when made of cast iron. Early crossings had a movable tongue that had to be set and held by hand for the desired route. The crossing is accompanied by guard rails that press against the backs of the opposite wheels to prevent the flange from striking the point of the frog. Cast-iron frogs were replaced by frogs built up of steel rail, which proved much more satisfactory, and eventually steel castings were used.

The second part is the switch, which actually selects the route. The earliest devices used movable switch rails that were individually positioned to align with the desired track. The pivot was at the heel, and the movable end was the toe of the switch rails. Later, the switch rails were connected and moved simultaneously by a suitable mechanism. This kind of switch was later called a stub switch. An alternative used a single movable tongue that could be placed to divert the flange of a wheel that would then pass on one side or another of a point on the opposite rail. Such turnouts were later used on street railways, in an improved form. A more reliable switch had one movable tongue on each rail. These points were connected to move simultaneously. This was called a split switch.

Each of the three possibilities had its own strong points and deficiencies. The split switch could safely be run through in the direction from the crossing to the point, called the trailing direction, perhaps with damage to the switch mechanism, but without danger of derailment. On the other hand, stones or ice could prevent the points from fitting properly against the stock rails, with the danger of derailment or improper route in the facing direction. Safety depended on the proper support of the thin point rails, which was provided by fitting against the stock rails. Stub switches, on the other hand, were largely immune to rubbish or ice, and did not involve thin point rails. However, derailment was certain if they were trailed in the wrong direction, and the spacing between the ends of the switch rails and the stock rails had to be large to avoid jamming. They were improved by making the movable rails the ends of rails unspiked for a distance and supported by cast-iron sliding chairs. A great convenience with stub switches was that the making of a three-way turnout was simple. In fact, photographs show that three-way switches were used wherever possible. The switch stand was the familiar "harp" with a lever that could be fastened in any of three positions, and would make the position of the switch evident.

In Britain, the nearly universal double track lines made it possible to use split switches only in the trailing direction on main lines, with only the rare facing switch under immediate supervision where necessary. In America, the prevalence of single track made this solution impossible. The more severe weather with its icing problems led to greater acceptance of the stub switch, especially in the west, but both kinds were widely used. Turnouts were treated with caution, and severe speed restrictions were placed on facing switches. High switch targets were used to make the postion of a turnout visible at greater distances, and when fixed signals were introduced, these signals were often used to protect main-line turnouts. Still, split switches were often operated by a simple weighted lever that moved transversely, and could be padlocked in either direction.

The Pennsylvania Railroad preferred stub switches because of the problems with ice, but was not satisfied with them, though they were widely used. A turnout was desired that could be trailed safely when set improperly; these were called safety switches. The fact that the main line rails were interrupted in either the stub or split switch was another disadvantage. For the sake of some minor connection, the integrity of the main line was compromised. These desires led to the invention of several kinds of "safety switch" in which the pressure of the flanges against points in the wrong direction threw the switch to the correct position. This was usually accomplished by gearing or springs.

The Wharton switch was introduced in the 1870's on the Pennsylvania Railroad as a switch that did not require that the main-line rails be interrupted in any way. The idea of the switch is shown in the diagram below, which has been simplified to show only those parts necessary for understanding how it worked.

The main line rails AA and BB are seen to be continuous through the switch, which is shown set for the main line. The movable rails are GH and KH', one with a pointed toe and the other blunt. Typically, they were 18' long, and connected with fishplates at the heeels. When the weight L is moved to the other position, G and K are pressed against the stock rails, and the switch is set for the divergence. From G or K to a distance shown by C, 3' 6", the switch rails rose 1-3/4", a rise of 1 in 24, and were then level as far as their heels. Both the point and the blunt rails rose equally. This rise was not objectionable, since the speed was low on the diverging route, and the rise was experienced by only one truck at a time. Initially, there was a rise of 2-1/2" in 2', which did prove too abrupt. The guard rail E, which was rigidly attached to the switch rail, kept the wheel set over so that the flange of the opposite wheel did not touch the stock rail BB as the wheel was raised while the outer part of its tread rode on the switch rail KH'. When this wheel was raised the full amount, the flange then moved across BB and assumed its normal position on the diverging route. Later, the switch rail GH was rolled in one piece with the guard rail.

The problem now arose of making the Wharton switch a safety switch. Note that when the switch is open, the movable guard rail F, pivoted at M, is brought against the rail BB of the main line. Should a vehicle approach in the trailing direction on the main line when the switch is so set, its flange wedges the movable rail away from BB and throws the switch for the main line. A vehicle approaching from the divergence when the switch is set for the main line is more difficult to handle. In the normal course of events, it would simply run off the rails as at a stub switch. Additional fixed guard rails are provided beyond G and K that catch the wheels and guide them over the main line rails while still elevated, then allow them to drop safely. These rails are not shown in the diagram above. Originally, iron castings were used for this purpose, but the drop was too abrupt.

A cut from the Wharton catalog shows the switch layout. When the switch is set for the main line ("open" in Wharton terminology, but normally called "closed") the main line rails are continuous and uninterrupted. Note the pivoted "trip rail" that will throw the switch in case a train approaches from the heel direction. It should be clear how the outside of the wheel tread rides up on the inclined rail until the flange can pass safely over the stock rail. The single point guides the other flange, moving the wheel to the right.

A Wharton derail is shown in the approach to a junction at the left. Unfortunately, no signals are evident, but the signal tower looks like Jersey Central or Reading, and the date is around 1900. The train shown is taking the divergence, but the derail is on, as if the route on this track was obstructed. There is probably a point indicator looking like a high switch stand out of view to the right, since one is present for the derail in the other direction.

The Wharton Railroad Switch Co. was at 125 S. 4th Street, Philadelphia. In 1886, a Wharton switch cost from $70 to $100, with $40 extra if you wanted the safety rails. A target to show the position of the switch was $10 extra. The switch was not used in main-line junctions and crossovers, but for the numerous other connections to the main lines used principally by freight traffic and at low speeds, or for derails, as the cuts show.


J. C. Trautwine, The Civil Engineer's Pocket Book, 13th ed. (New York: John Wiley and Sons, 1888), pp. 778-779.

The cuts from the Wharton catalog were provided by L. G. Lovejoy, P.E., of Pittsburgh.

Return to Railway Index

Composed by J. B. Calvert
Created 12 October 2000
Last revised 21 October 2000