Article From the Summer 2001 Issue of


The Journal of the New York Museum of Transportation


In the spring issue of HEADEND, we described how the chance discovery of an old newspaper article piqued our interest in experiments designed to use compressed air to power streetcars, before the advent of commercially viable electricity. In Part I, we discussed early ideas for moving messages and people down air tubes, with many systems built and operated as early as the mid-1800’s. For more technologically advanced concepts, read on:

Michael Wares, at Fordham University, reminds us that William D. Middleton’s book "Time of the Trolley" briefly discusses some attempts to harness compressed air in transit systems. The Third Avenue Elevated, in 1881, experimented with something called the Hardie air locomotive, in an attempt to reduce air pollution and danger from sparks from their steam locomotives. With four 120 cubic foot air reservoirs where the boiler would ordinarily be, the device was charged to 600 pounds per square inch at the beginning of its run, and was even designed to use its cylinders as air pumps for braking (although it’s unclear whether this helped recharge the tanks in an early form of regenerative braking). Wares also says that due to the prohibition of overhead trolley wires in most of Manhattan, compressed air (and storage battery) cars were tried there too.

We also know that, just as some industries used "fireless" steam locomotives (using steam from a central plant), there were others, such as coal mines, that employed compressed air locomotives. As recent an example as the Locomotive Cyclopedia for 1941 shows an H. K. Porter model operating at 250 psi, from a tank charged to 800 psi, developing 8,000 pounds of tractive force.

Mark Brader tells us that the French city of Nantes (where, in 1826, the horse-drawn omnibus was first used) developed a large fleet of compressed air streetcars that started operation in 1879. Mark says that a model the line introduced in 1900 had an under-floor storage tank of 95 cubic feet capacity and was charged to 80 atmospheres of pressure (1,200 psi). He points out one of the shortcomings of compressed air systems: the air had to be heated before use, as cooling from the expansion of the air caused ice to form in and on the cylinders. Other technical hurdles would have been the need to periodically recharge the car’s tank and the relative inefficiency of compressing air (see box). One might wonder, too, about sitting over a 1,200 psi "bomb". Apparently it worked for the French, as the concept was in use until 1917.

Shelden King, in his book, "New York State Railways", cites the March 1927 issue of Transportation News (the employee monthly of N.Y.S. Rys.) as he describes Rome, New York’s brief and rather late flirtation with compressed air. Horsecars were first operated in Rome on June 27, 1887, and they provided service until well after most other cities had converted to electricity. The American Air Power Company bought a controlling interest in the line, and on September 22, 1900 the first run was made of a car powered by a Hardie air motor.

According to the Transportation News article, "A car started out in the morning with its tanks fully charged and when the air pressure became low, the motorman returned to the power house to have the tanks of his car recharged". The article doesn’t say how often these recharging trips had to take place, but does say "it was often necessary to send out to get a car which had run out of air upon the road". From mid-1902 until the inauguration of electric trolley service on June 9, 1903, horses did most of the work on the Rome system.

Here in Rochester, we find further information in "The Genesee River and Its Relation to the Rochester Gas and Electric Corporation", written by RG&E employees and published in 1943. Tom McColloch, Senior Mechanical Engineer in Fossil/Hydro Administration for RG&E, has graciously provided us with selected portions of this book which offers further illumination. According to the book and its appendix listing several newspaper articles for further research, James M. Bois, born in nearby Aurora, NY in 1842, invented the Bois Compressed Air Motor and formed the Hydraulic Motor Company. In a May 13, 1880 article in the Rochester Daily Union & Advertiser, work on the company’s facility at the Lower Falls was said to be ready to commence in a few days. A subsequent January 5, 1881 article in the same paper reports on an inspection tour and somewhat opaquely attempts to explain how the system of vertical and horizontal chambers was intended to work. Apparently the system employed static water pressure, as only about 45 psi was expected from the 97-foot head.

The articles about the Bois system are enthusiastic about replacing horsecars with compressed air machinery. Horses (and mules) could only stand a few hours of service in the best conditions, so even a system the size of Rochester’s required hundreds of animals, with the attendant costs of feed and maintenance, not to mention the immense manure problem. The equine epidemic of 1872 all but wiped out the stables of many horsecar lines.

Bois apparently had some bugs to iron out of his system, as we next hear from him almost two years later in a November 24, 1882 article in the Rochester Daily Union & Advertiser (researched for us by Charlie Lowe). Bois is characterized in the article as having been "laboring long and hard to prove the practicability of his idea". The article describes an actual demonstration on November 23, made on tracks laid "half way from the car house at the bank of the river, just below the falls, to the tracks of the [horse-powered] street car railroad company". Robert Hardy [sic], "the inventor of the pneumatic engine recently introduced on the New York elevated railway", was at the controls as the noisy car sped back and forth on the short section of track. The Bois plant was now capable of providing compressed air to street level at 75 psi, but statements were made at the demonstration that at 400 psi, the car should be capable of carrying 30 passengers over 10 ˝ miles. This seems a reasonable claim, as the previously mentioned Hardie air locomotive tested on
This c. 1882 photo of the Lower Falls shows Bois’ compressed             the Third Avenue Elevated back in 1881 powered a
air plant in the foreground. From "The Genesee River and Its              three-car train for 9 miles.
Relation to the Rochester Gas and Electric Corporation",
published By RG&E

Indeed, just a year after that first Rochester demonstration, a much larger-scale demo took place. The November 7, 1883 Rochester Daily Union & Advertiser (also provided by Charlie Lowe) included a brief note that "A street car on the North St. Paul Street route, yesterday, was run by the compressed air motor and worked very satisfactorily. Over thirty trips were made and 1,000 passengers carried, the run being from the Central depot to the Asylum".

But the compressed air crowd wasn’t the only game in town. Among others interested in harnessing the power of the Lower Falls was the Brush Electric Company. According to "A Century of Engineering in Rochester", published in 1997 by the Rochester Engineering Society, Brush built a generating plant there in 1881 just after Bois’ compressed air plant was erected nearby, and on October 29 of that year the first lights for commercial use were turned on in the Powers Art Gallery and the A. S. Main Store in the Powers Block on Main Street in downtown Rochester.

After all this, what we’re still not sure of is just how Mr. Bois’ concept worked. Newspaper accounts obviously were written by non-technical people and are not too clear. We hope it didn’t involve too much machinery as Canadian Charles H. Taylor invented a system of compressing air using water power and no moving parts at all! In the Ragged Chute plant built in 1910 and still in operation, water from the Montreal River drops 350 feet down a vertical shaft and makes a sharp turn to a horizontal tunnel at the bottom. Air in the water compresses in the fall, but separates naturally from the water in the tunnel from where it is piped off to local cobalt mines. The system develops pressure of 862.5 kPa (125 psi).

Of course, what we do know is that the first electric trolley operation in the country got started in Richmond, Virginia in 1887, and—except in Rome, NY—any thoughts of further local experiments with compressed air machines were probably discarded in the race to join the electric age.

As we embark on the electric age ourselves at NYMT, it seems appropriate to reflect on the birth of electric power and the trolleys that eventually became commonplace on city streets and then faded into history. Imagine all the alternative scenarios that could have developed into reality—mechanical transmission lines from Niagara Falls, compressed air subway cars, and pneumatic tubes instead of e-mail. Many of today’s cities are still exploring alternatives, and as a result, the modern light-rail equivalents of streetcars and interurban trolleys are flourishing around the country. The compressed air experiment on St. Paul Street happened over a century ago, but the lesson is as modern as tomorrow.

We thank a number of friends of the museum for taking the time to help us research this article: Mark Brader, Michael Wares, Shelden King, Tom McColloch, Charlie Lowe, and David Minor. You can listen to David’s Time Traveler series each Saturday morning on WXXI-FM (91.5). Check out Mark Brader’s entertaining review of urban transit development on the web at

There is also a site devoted to compressed air operated trolleys and locomotives that might be of interest,

What do you think?

We wondered about the numbers associated with vehicles powered by compressed air machines. How fast would they go and how far before needing a time-consuming new charge of compressed air? Beginning with the definition of horsepower as 33,000 ft-lb/minute, and assuming 10 horsepower for a machine to run a streetcar with any degree of zip, we started to do some back-of-the-envelope figuring. We assumed some cylinder bore and stroke combinations, gear ratios, wheel diameters, and tried a 3 ft diameter, 5 ft long tank with air initially at 300 pounds per square inch.

We ended up with a streetcar that would go 7 miles per hour but only had enough air to go about two thousand feet down the track. No wonder Nantes used 1,200 psi!

Maybe we made a mistake in our calculations, and/or maybe you’d like to try arranging the variables yourself to produce a design that would satisfy the traveling public. Let us know what you come up with! If it’s a really good analysis, we’ll let you play the calliope the next time we crank ‘er up. Okay, sharpen your pencils! Ready, set….