Conventional wisdom has it that the last time a steam engine headed a regular San Francisco "Peninsula Commute" train was on January 22, 1957. This also occasioned Southern Pacific's last use of a standard gauge steam locomotive in regular service in the United States (its Mexican subsidiary killed its fires at a later date).

The second statement is true, but the first one is patently false.

When I went on duty as a Caltrain* engineer on the afternoon of December 16, 1994, it looked to be like most days. I would run a train from San Jose to "The City"and return, mostly ho-hum-boring assembly-line work, perhaps with an off-chance moment of holy terror. (*Caltrain was/is the government-owned successor to SP's Peninsula Commutes that contracted-out the operation to Amtrak for operation. Thus, I actually was an Amtrak employee.)

AH, but blind luck, perhaps aided by a deity or two, reckoned otherwise. This would be far from a ho-hum or terrifying day - quite the opposite. As my train swayed through the crossovers that formed the yard throat at the San Francisco depot, w-a-y over to my right a sight came into view that made my jaw drop: the Golden Gate Railroad Museum's pride and joy, ex-SP P-8 Class 4-6-2 #2472, WAS BACKING ON TOP OF MY TRAIN, #74!

HOLY COW! I was enough of a historian to recognize that this was a unique moment, the first time in nearly 38 years that steam had pulled a regular Commute train on the San Francisco Peninsula! As it has sorted out thus far (2014), this also was the last time.

After I tied down my engine, I grabbed my "orders" and nearly ran to #74 to see what was up. Caltrain engineer Bill Stetler, who had trained on the 2472 under the tutelage of veteran SP Engineer Neil Vodden, was up in the cab, fiddling with appliances about whose functions I could only guess. Bill had come to Caltrain a few months earlier, and this turned-out to be the first of many 2472 runs that we would make together. Also in attendance was Amtrak "Special Duty" (whose nominal job description was that of a trainmaster) officer Bob Bongiorno, another SP steam man who hired out in 1952 and whose railroad career would last fifty years, to the day.

Bob gave me, the official "engineer of record" (Bill was an unpaid volunteer) dispensation to do what I pleased, since the higher-up's contemplated us using the train's F40 diesel merely as a supplier of head end power for the train's electrical needs, with its engineer's controls cut-out. I was free to ride the cushions back in the cars, or ride the 2472.

You bet, the first thing that crossed my mind was to climb onto the 2472 and ride it all of the way to San Jose! What railfan would not give a right arm and leg for such an opportunity? Myself, I had occupied the cab of a steam locomotive for a long distance only once in my middle-aged life, eons before. In terms of steam knowledge and experience, I was on the opposite end of the knowledge spectrum from Bill and Bob, and probably many railfans, as well.

In 1954, apparently backing up to a joint on #144's equipment, 4-6-2 #2472 shares the SP's Third & Townsend depot with 4-8-2 #4320 (left) and larger 4-6-2 cousin, #2487.

- B&W slide from Wx4 collection
photographer unknown

But, my second thought was a more-conditioned one. This was a commute train, not a foamer flyer. The passengers had worked all day, were tired and had a rightful expectation that we would have them home on time, before their dinners got cold. I was being paid to hold-up my end of the bargain. Railfanning at the expense of the passengers was simply unconscionable. I understood the general situation, that working alone, the 2472 did not have any hope of staying on time. Number 74 was a fairly heavily-patronized train due to leave at 7:00pm, as the first makes-almost-all-stops local after the Fleet (rush hour express trains). The weight of that F40 and five cars was a lot for the 2472 to handle, and Bill, nor anyone, had operated a steam locomotive on "The Commutes" since 1957. Heck, except for a paycheck, about all that a commute engineer had to show for his efforts at the end of the day is the fact that he/she got the folks home safely…and hopefully on time.

I asked Bob if there would be any problem in running the train as a double-header, explaining that I figured that I had enough background running helper units in Dunsmuir to enable me to help Bill more than I hurt him. He replied that it was OK by him, but that the matter should ultimately decided by Bill. When I climbed up to the cab, it was apparent that Bill was way ahead of me on all of this, and we quickly agreed that I would give him a boost out of stations. I told him that I would throttle-off at 50 mph, or when the needle "wiggled" on the brakepipe gauge (i.e., he applied the brakes), and that he was otherwise on his own, especially while braking. In truth, I could not help much in the latter regard, anyway, since Caltrain F40's did not yet have dynamic brakes, and Caltrain was having overheating problems with the F40's wheels, which at the time we wrongly guessed was caused by overuse of the independent (engine) brakes. This made the "jam" (independent) off limits.

After syncing with Bill, I went back to prepare my loco, F40 #916. I anticipated having a railfan's catbird seat, which would only be diminished by the fact that the 916 was what we called a "growler", a locomotive whose prime mover (engine) constantly ran in Run 8 (maximum throttle) to supply electricity to the train. I later found, however, that if I stuck my head out the window, I could hear 2472 bark out of stations just fine.

It all felt surreal as I sat there in the cab in the dark. In the rear-view mirror, I watched the passengers walk along the platform hunched-over against the cold and climb aboard the cars without any notice of that huge cloud of swirling steam on the head end. I only saw one guy take notice and walk up to the 2472. The reason that the steam engine went so unnoticed was that the number two man on Caltrain's staff, Walt Stringer, had decided, in a master-stroke of public relations ineptitude, to add the locomotive to the train at the last moment, apparently not bothering to tell anyone other than Amtrak operating personnel. The occasion for this ferry move was the next day's planned Shoppers Specials (the precursor to later Toys for Tots Trains), which the 2472 was slated to power. (I worked the first leg of that round trip, as well)

2472's career:
Built by Baldwin and placed in service in 1921, along with 14 sisters, #2472 spent its early life as prime power on SP's Overland Route out of Ogden. By 1940, she was assigned to the Coast Division, where she handled a variety of freight and passenger assignments. The loco mostly worked the Peninsula Commutes, an assignment that she performed up until the end of regular steam operations. She was vacated in San Jose from the roster on February 7, 1957, about two weeks after regular steam on the SP's Commutes came to an end.

After retirement, she was placed in the deadlines at Bayshore Yard, San Francisco, but escaped scrapping long enough to be donated to San Mateo County in 1959. SP gave her a coat of fresh paint and on April 10, hauled her to a display site at the San Mateo County Fairgrounds, where she prominently sat for many years. The message at left describes that move.

In 1975, a group that would eventually become the Golden Gate Railroad Museum began restoration efforts on 2472, which were not completed until 1991 at Hunter's Point Naval Shipyard. During her early runs, SP engineer Neil Vodden trained Bill Stetler on the locomotive's operation. Bill was a kid when the restoration originally began, but nevertheless was the engineer on most of 2472's Caltrain runs in the 1990's. The later abandonment of San Francisco's Mission Bay Wye (now the site of a University of California campus), where 2472 turned around for southward trips, was a severe blow to 2472's continued operation on Caltrain.

Sadly, although the loco still is in prime shape, it currently sits cold in storage, with nowhere to run upon since the 2472 and Niles Canyon folks parted ways, and Union Pacific banned steam locomotives, other than its own, from its lines.

When 7:00pm came, we highballed out of town to the notice of few, except that it really wasn't much of a highball, since various speed limitations kept the train at a relative crawl until the 22nd Street depot. Bill was able to open up his engine departing from there, and with my extra 3000 horses, we rocketed up to the track-legal 50 mph in no time. At Bayshore, the second stop, the track speed was 60, so Bill was able to open her up even more. All that I did besides run my loco was utter "WOW!" to myself.

At South San Francisco, the third stop, I noticed both the first railfan watching the show and that we were only about a minute behind schedule. This actually was a little better than normal for Train 74. After the next stop, San Bruno, Bill was able to take his engine up to 70 mph, and by then I was appreciating what a prodigy he was in controlling it.

Running a commute train is a very busy endeavor - the engineer is always doing something in-between those stops of every two-to-four minutes. Along with all of the normal duties involved in running a diesel-powered commute train, steam engineers (so I later observed) work in an arrangement of controls and cab appliances that only vaguely resembles that of a diesel. The overall difference in the cab environment is dramatic, as well. When I later rode the 2472 at high speed, I was amazed at how physically punishing the trip was. The open cab was frigid near the windows and hellaciously hot otherwise. The ride jarred my bones into unnatural relationships with each other. My field of vision ahead was partially obscured by that long boiler, which was problematic, because the noise was so volcanically deafening that shouting at the top of my lungs directly into the ear of a cab-mate was the only way to express complex thoughts. Running diesels was not exactly exactly like propping one's legs on a sleeping pet fido before the warm glow of a living room fireplace, but it might as well have been, in comparison to operating a steam loco.

I imagine that a reasonable way to explain to a non-engineer the overall professional challenge that Bill faced that night is thus: Say that you've spent your normal weekday life for the past ten years commuting in your trusty old Subaru in rush hour traffic, and on sporadic weekends you practiced driving a Kenworth semi tractor-trailer on uninhabited country roads. Then suddenly, your Subaru quits and you're forced to commute in your Kenworth, and incidentally make sure that you get to work in time for that morning meeting. Remember, the 2472 was coupled on #74 at last-minute's notice, and Bill was forced to adjust to the situation on the fly.

The obvious danger, to you, the Kenworth driver, is that the novelty of the experience will cause you to fixate upon the actual operation of your truck, thereby overlooking the fact that you are about to plow into a school bus. For engineers, focusing upon, and adjusting to, the ever-changing conditions ahead of the train is paramount. The hardest-won skills are the ability to treat the actual physical operation of a locomotive, as well as deal with a multitude of distractions, as autonomic processes that operate independently of your decision-making. In addition, imagine trying to keep a Kenworth sans a steering wheel from hitting that school bus, while using only the truck's parking brake. Under emergency braking in favorable conditions, a commute train can take up to a half mile to stop from 70 mph. At that speed, my rule of thumb about braking from high speed for unexpected occupants ahead is this: If I see you, I will hit you, unless one of us gets off the tracks.

Under "controlled" (normal) braking, stopping distance can push three-quarters of a mile, and even then, it takes considerable amount of practice to stop at a precise point, (often within a window of three feet, or less) because, again, all of that inertia just wants to keep on going and going like the Energizer Bunny, particularly if the brakes are not functioning optimally, which is often the case. There can be a wide disparity between "optimal" and "legal functional". Given all day and enough distance, anyone familiar with the controls could crawl into a stop in reasonably good order. But commute train schedules are based on quick, efficient stops. To accomplish this, engineers often enter stations using every last bit of braking effort available (in the vernacular: "throwing out the anchor", or "platform diving"). The thing is, the difference in the actual physical sensation between moderate and full braking is very slight, and the braking characteristics of two identical-looking trains can be disconcertingly different. What this means is that it is easy to get fooled and put a car, or two into the "rocks" (ballast) beyond a station platform - a big no-no. I certainly was not exempt from this.

So, getting back to Bill, he was faced operating a train knowing full-well that he faced a heady combination of the known and unknown. He had previously run diesel engines on Caltrain for several months, as well as occasionally operated the 2472 on excursions, BUT he had not previously put the two together. He could expect a quick takeoff due to two working engines, but likewise these engines could not contribute much towards stopping the train. The diesel's aforementioned issue with brake-overheated wheels was serious enough, but the steam engine drivers' sensitivity to heat was much worse. If Bill rode the 2472's engine brakes too much, he could slip a tire off of a driving wheel. Back in the late 1940's an SP engineer did just that, put a mail train into a Salinas Valley ditch by "riding the jam".

As I earlier said, by San Bruno it had become apparent to me that Bill already had mastered the situation, especially the stops, which were flawless. It all went so smoothly thereafter that I cannot remember any subsequent outstanding moments. I just boosted him to 50 out of the stations, then throttled back to Run 1 and enjoyed the show. Repeatedly, I consciously said to myself, "They're paying me for this!" in addition to "WOW!"

We arrived in San Jose about ten minutes late, about normal for this train, based upon my experience. So, I judge that the extra acceleration provided by the diesel and the additional time lost in braking for stations (due to all of that extra, largely un-retarded engine weight) effectively cancelled each other out.

Thus, Bill put us into town just like it was any other workday…back during an era before he was born. I realized that, given the same training and experience, I could not approach the performance of a guy who was a decade and one-half younger than me! I was, and am, in awe.

The 2472 steams through South San Francisco in 2003, several years after the balloon track episode. Photo courtesy of former Caltrain Engineer Frank Caron.


Bill left Caltrain only a couple of weeks later to go to work for Burlington Northern, during one of Amtrak's more-severe annual budgetary/existential crises. Much later, his considerable skills led him to become the head of Canadian Pacific's steam program. He returned to run the 2472 on many occasions, and on a fair number of them I was either behind him on the diesel, or with him in the cab. Here's two interesting subsequent incidents:

The first took place the next year, I think, when Caltrain again ran the steamer on Christmas specials. During the evening up-trip to San Francisco, just before we began to pull out of Menlo Park, I noted a reasonably well-dressed middle-aged guy standing a little too close to the 2472 for comfort, as he quixotically put on a pair of gloves. When we began to move, much to my horror, he grabbed ahold of a pipe AND HOISTED HIMSELF UP ONTO THE SIDEROD that connected the rotating drivers! Up and down and up and down! My first thought was to plug (apply the emergency brake) the train, but to do so at slow speed might slam passengers to the floor. The nutcase was below Bill's field of vision, so all that I could do was yell into the radio microphone, "2472, that'll do, Bill, that'll do!" Unfortunately, Bill was working with a new cab radio arrangement that was not adequate to the task of being heard above the thunderous cab noise). No response. Finally, on the third try, he barely heard me and brought the train to a rapid, but smooth, stop. The guy jumped off, and disappeared running down Oak Grove Avenue. This certainly gave new meaning to the old hobo term of "riding the rods".

The other one happened later, in the late 90's, I think. Previously, 2472 turned around in San Francisco as in the past, on the Mission Bay Wye, near the long-gone passenger steam locomotive roundhouse. On that day, the wye was out of service, so somebody (perhaps Bob Bongiorno) suggested that that we try turning the loco on the former South San Francisco Belt Railroad balloon track in its namesake city. A balloon track, as the name implies, is a circular track whose intersecting switch is so arranged that a movement entering the loop winds up exiting facing in the opposite direction.

The balloon had not been use for several years, and the curves were pretty tight - I doubt that a locomotive as large as the 2472 had ever been sent around it before. The UP sent their section crew to aid in the move by clearing rubbish and oiling switches, but as it turned out, we needed them more than anyone had figured upon. The first two or three hundred feet went OK, but then the UP track foreman abruptly yelled for Bill to stop, even though I doubt we were doing even one mph. The lead drivers were climbing the rail, due to some particularly sharp track curvature. Bill then eased the engine back a few yards so that the UP gang could widen the gauge (distance) between the rails enough to allow the engine to pass. Bill eased through the widened section, only to be stopped again on the other side, this time because the center drivers had jumped the rail. This sort of thing happened two or three more times, complicated by the fact that everyone on the ground was busy behaving like a sidewalk superintendent (me: also guilty). I generally knew/know nothing about steam locos, but I was nevertheless wondered what effects that this harsh handling was having on the 2472's fabricated frame.

None of 2472's wheels ever fell to the ground, thank goodness, and we made it around the balloon after about an hour-and-a-half of fitful progress. Although the exercise apparently caused the locomotive no ill effects, they never tried THAT again.

THE Photo

During 2472's run in #74, I did not notice a single flash bulb go off, likewise spied no photographers at all. So, for almost exactly 22 years, I was of the belief that the historic moment had eluded visual record.

A few days prior to the mid-December, 2014 day that I am currently typing these words, I posted a much-abbreviated version of the above story on, a railfan discussion site, in the outside hope that somebody, somewhere possessed a picture of the momentous event.

Sure enough, somebody did! A gal who only wishes to be known as "Margaret (SP Fan)" had slipped by me on the San Francisco platform and had managed to take the picture seen at left. That's Bill Stetler in his trademark overalls, along with an unidentified individual.

Margaret got involved in the 2472's restoration program early-on, and spent time cleaning the inside of the boiler and performing other grunt work. She has been described by a fellow as "for many years the heart and soul of the group and a great volunteer", and I can personally confirm that her enthusiasm for railroading extended to we appreciative employees.

So, what you see here is what I regard as a Christmas miracle.

Margaret (SP Fan), I shall always treasure this snippet of one of my favorite moments in my career, and I love you dearly for your warm generosity!

-E.O. Gibson, engineer emeritus

SP Index