If so, how did they work?
Hi Chip,
I believe that they did use towers to control interlocking plants and signals that early, perhaps earlier. This was done by mechanical means and those parts often didn’t change much until into the 1950s-60s, although that new fangled electric stuff starting getting hooked up to it eventually.
If you look at pictures of towers, you will usually see various arrays of rods placed on low foundations that run from the tower to the signals and turnouts that it controls. The tower operator tugged on big levers to line things up for each train. The telegraph was a part of this, since it communicated changes in schedule to the operator and thus the authority that would be communicated by the tower operators as they dealt with each train.
I have a little info I can pass along. I know there were interlocking towers then, and they operated in much the same fashion for most of their usage. Here’s some examples that I know of from that time period.
http://en.wikipedia.org/wiki/Interlocking_tower
dlw-sussexbranch.com/Stations/waterloo.html
arrts-arrchives.com/atlaverr3.html
user.mc.net/~louisvw/depot/clare/clare.htm
Here’s a bit on how interlocking works, at least on the old NY Subway system.
As an extremely broad answer… based on recall from the research i did…
Towers existed in the US in the C19 where there were numbers of lines and large/larger amounts of train movements to keep apart/regulate. In other words where the traffic paid for a tower. Also, regulation of train movements is a factor where train movements are close together physycally and in time, higher speed is required and movements can conflict.
The NE and big cities were the first to invest in interlocking /towers. Way out in the wilds frequently never justified it and never got it.
If your trains didn’t need to keep rolling to keep out of the way of each other - and labour was relatively cheap - you could have local crew or a train crew that would unlock and move switches as and when needed.
Once you move into the C20 you start to get remote interlocking to an increasing degree… but that’s not what you are after.
From what I know of your layout plans so far a tower is not likely to be appropriate but ground throws and trains stopping (or crawling) to let crew off to throw them would be rght.
[8D]
Dave’s right - the one spot you would be likely to find an interlocking in the West in that period was at a crossing between two lines. There were some very odd ones - the only signal on the whole of SP’s Modoc Line was at Stronghold, CA, where the GN “Inner Gateway” connection to the WP crossed the SP. There was also on near San Luis Obispo where the SP’s Coast Line crossed the narrow gauge Pacific Coast Ry. The PC was there first, so the responsibility for building and operating the signals fell on the SP. The traffic disparity between the two was huge and got bigger until the PC was abandoned in the 1940s - they eventually negotiated an agreement so that the route would be set to keep the Coast Line clear, unless a PC train showed up. Toward the end the PC service was so irregular that the SP didn’t even station a signalman in the interlocking unless the PC notified them beforehand that they were going to run a train, and one of PC’s last excursions had to wait at the signal while someone went to hunt him down so he could clear a route through the diamond.
But both Reading and C&O built very attractive towers at that time- C&O’s were octogonal, and Reading’s were an odd, tapering style nicknamed “windmills.”
The short answer to your second question is, as Dave noted, “mechanical interlocking.” Dave has posted a lot about British practice on the discussion we had on modeling UK prototypes, and there are some books that discuss American practice - there’s an excellent essay in Thomas Clarke’s 1887 collection of essays called “The American Railway.” There was a discussion of this book on the EarlyRail group awhile back - someone may have posted text online; I got a reprint in 1976 (one of my first railroad books) that I still use; the illustrations are superb. There’s a picture of a Reading tower in it.
Chip,
The Pennsylvania Railroad used towers as block stations in the 1880s every few miles.
Dave,
Did they actually control entire blocks/intervals, or just interlocked crossovers? I was discussing this with Dave on another thread, and I thought they used timetable and train order, with the interlockings at regular intervals in contact with the dispatchers to control interlocked sidings and crossovers.
As far as I know, when they were first installed on the PRR main, they controlled entire blocks as well as the interlocked crossovers. Evetually, as the PRR migrated to CTC, they just controlled interlockings.
I’m less confident in their usage as I am about photographic evidence that at least they existed in the 1880s along the Pennsylvania Philly to Pitsburgh main.
I know Messer’s book on the Middle Division shows them at frequent intervals, and shows most of the interlocking diagrams, but he doesn’t talk much about operation. The B&O used a similar setup on their Cumberland Division, which was also a pretty high-density operation.
The best explanation I have found of how they did this was in the NTSB’s report on the accident at Chase, MD in 1987 - you can find it online, and it has a lot of stuff about signals and procedure. It’s not Pennsy practice per se, but my guess is that Amtrak inherited Pennsy procedure with its plant and staff when they took over the NEC in 1977. The NEC was being signalled for CTC when the wreck happened, but it still had interlockings operated independently of the dispatchers in Philly - I think the intervals between interlockings were signalled with automatic block signals. It sounded like the dispatchers talked to the block operators, and gave them instructions on routing - for instance, set the switches and signals so that this train will move north on that line; the dispatchers kept track of trains on train sheets - they still gave train orders, apparently (one of the first things the dispatcher said after realizing that the accident had happened was “I better start getting out some hold orders”), but they must have relayed them by radio. It sounds like an intermediate step between timetable-and-train-order and full-on CTC.
Which Dave? [(-D]
In general US practice the answer would appear to be “Yes… and No”.
In dense traffic areas the US roads appear to have adopted mainly UK systems of Absolute Block Working. I was going through some files the other night and one eample that caught my eye was Patenall’s adaption of Syke’s “Lock and Block”. Lock and Block is a rotational locking that ties up the Block Instruments with the signals in a sequence that must be completed or keyed out of according to the Rule. (Anyone with “Red for Danger” LTC Rolt look at the South Croydon smash - this was on my old patch - but before my time). This is the Yes part of the answer. Full Absolute Block Signalling is intensive in equipment and man power and therefore epensive to install, maintain and operate.
In non-dense operating areas an interlocking tower just interlocks what the company wants t to interlock. This may be a diamond, a junction or multiple lines. What the interlocking is doing is bringing all the controls (that are selected0 together where one (or a very few) operators can control them both with an over view of what is going on and the protetcion of the interlocking machinery within the frame to prevent them from setting up opposing or conflicting routes.
This is where the simplest interlockings come in. At a diamond with enough traffic what signals there are are interlocked so that only one route across the diamond can be cleared at one time. This may be in both directions if there are two roads.
As has been rightly
That has to be a rules violation of some kind! Actually, I was asking Dave Vollmer if he knew anything about Pennsy-specific stuff, but you’re obviously welcome here, too - you gave Chip a much more detailed explication about signalling than I did. Have you found out much about Absolute Block working in the US? I ask because in looking through Messer’s book about the PRR’s Philadelphia Terminal (a logical site for that if ever there was one), I haven’t found anything about it - it looks as if it’s automatic block between interlocking towers that control junctions and crossovers. The Reading is another plausible candidate, too.
Was the South Croydon wreck the one where there was a small flaw in the interlocking logic that allowed a lever to be thrown? That’s an excellent book - a lot of the reports Rolt used to write it are online now:
Another aspect to consider is “Train Order” working and Timetable working.
Taken the other way round…
Timetable working.
RR would publish a schedule of trains to operate over their tracks. The Timetable. Unless overruled by an appointed officer (such as a Dispatcher) this was not just the times trains should run but the order in which they ran.
many TT established that one direction had precedence over the other. Under this an Eastbound with precedence would be allowed to run through a section of line before a westbound regardless of whether either train was running to time or not - unless overruled.
So an on time Westbound could be held for late running Eastbound to clear.
Train Orders…
These are the things that the Dispatcher or other Officer would use to over-ride the instructions of the published timetable. They could also over-ride previous Train Orders.
The system is basically that a central office has a diagram of where all tracks, sidings and everywhere a train can be placed. It has one form or another of telegraphic communication to a number of outlaying telegraph offices. If these telegraph posts are designated they are “Train Order Stations”.
A Train Order Station is a telegraph equipt office that train crews know may give them instructions on the running of their trains.
Some Train Order Stations have no signal. In this case all trains will stop for instruction. More usually Train Order Stations had a “Train Order Signal”.
Train Order Signals took various forms but commonly had two semaphore arms on one post facing in opposite directions. The right hand arm usually applied to an approaching train. The Indications were usually:
Vertical - Clear/No Train Order to collect.
45 degrees - Train Order to collect on the move - by both loco crew and conductor. (With a multi-crewed loco lash-up the loco/crew designated as
Yes, and the earliest were made from Cornstalks.
That’s where the term ‘‘Cornfield meet’’ came from.
“Flimsies” were made of onionskin paper, very light, and designed for carbon copies - in pre-Xerox days, the op would write the order out with carbons for the additional copies (engineer, conductor, his records). Saved time and effort.
Don’t know which Rule but it’s why every conversation over radio or phone must start with a statement of who is speaking from where and that must be confirmed back - BOTH WAYS.
For example…
When I call a Signalling Centre I want to hear ’ “Name of the centre” Signaller speaking’. I will then say 'Hi “Name of centre” Signaller, this is “Me” speaking from xxxx…" and then carry on with the meassage. He then knows that I believe i am speaking to the right person…
Finding out about Block Working in the US is frustratingly difficult. I have so far failed to find a single general source. UK type Block Working only seems to have happened on any large scale in the North East. I can’t work out whether it was used much around Chicago or any other of the larger cities with complex and intense rail systems. It also seems to have been dropped or superceded quite quickly in the North East.
The head-on that you are probably refering to in which part of the interlocking apparatus deflected allowing a conflicting movement to be signalled was probably Hull Paragon. IIRC two “Achilles” Class locos head butted each other at fairly low speed. most of the damage was in leading frames.
South Croydon involved Sykes Lock and Block and the
Carbon was used less to save effort and more to prove that the copies were eactly the same… in particular the record copy had to be eactly the same as the other two - if there was a train smash. IIRC all orders were uniquely numbered to their location… possibly taken from an “order book”.
i don’t know whether the US used the practice of doube sided carbon paper. this puts a copy down to the sheet below and a mirror copy up to the sheet above… almost impossible to alter without it showing up.
[8D]
I thought that a “cornfield meet” came from lovers ‘just happening’ to meet each other out of sight in a corn field. I have a copy of a US woodcut of two trains piled up entitled “They met by chance, the usual way”.
[8D]
The only real train orders I have ever seen were in Ecuador, where they had apparently imported the US system verbatim - right down to the form numbers (17 and 19). But they came out of a pad, with the onionskin sheets in different colors, just as you have described. They were still using a telegraph and Morse to send messages - and this was in the 1990s.
Here is some infomation on old towers. I’m retired signal maintainer. One of my first towers was Davisville R.I. and Cranston R.I. The interlocking macnine was what we called a “armstrong machine”. Switches were conected via pipe line with rollers and cranks and operated by pulling a lever in the tower that moved the pipe hence the switch. An interlocking means that switches and signals had to be throwen in a certian order. Dogs in locking bed of the machine would slideand move to allow the machine to move the pipe.
A old timer told me that before track circuits They had a system where there was a pole that had a red ball on a wire that could be raised and lowered to signal a train. When ball was at top of the pole it ment track clear (high Ball, full speed a head). Raisad half way (look out for train ahead). At botton position ment Stop. Tower operator would talk towers ahead and behind via telegraph. When I started working we had a bell system from tower to tower. By pressing a old telegraph keep we would “O S” a train (on station) by the bell to let towers know where the train was.
Bell signals? Sounds like Dave might have found his American application of absolute block signaling.
What company did you work for - the New Haven?