History--Eight Digit US telephone numbers? [telecom]

I think I found examples of eight digit diable telephone numbers. Several ads in the NYT in 1947 had HOllis 5-10nnn numbers (HOllis 5- was in Queens, NY). Old Bell System literature says some large city exchanges could have as many as 10,500 lines--this was the maximum jacks within reach of an operator.

I was wondering how common this was in the US (presumably it was in big cities), and how long it lasted.

Some background:

The Bell System developed the 'panel' dial exchange for big cities. It was designed to have more capacity and switch more efficiently than was possible with step-by-step gear; it utilized some basic common control circuits.

A key feature of panel was compatibility with the many manual exchanges a switch would connect to. For calls from manual exchanges to a dial exchange, operators had a keypad to enter the number quickly (either the originating operator or dedicated "B" operators* for that function).

For calls from dial users to a manual exchange Bell wanted the connection to be easy from the subscriber's point of view. Thus, dial subscribers dialed all their city calls regardless of the type of exchange they were calling--they did not have to know whether the called exchange was dial or manual. In this way it was easier to convert manual exchanges to dial.

To accomplish this, inward operators at manual exchanges had a display panel indicated the desired number. The display was controlled by the dial switch, which translated the called number into a signal to light the appropriate signal lamps.

The literature on this notes that a manual exchange could have up to

10,500 lines (not merely 10,000). The display provided for this by having a leading fifth digit of 0 or 1. Presumably callers to such an exchange would dial eight digits.

(The inward operator display also accomodated the party line suffix letter, but that is beyond the scope of this discussion.)

So, would anyone know if eight digits were actually dialable in New York City or in other places, as the literature suggests? If so, how long did it last? I presume as preparations for DDD came along the eight digit subscribers got new numbers.


*In large cities, manual switching required two operators--an "A" and "B" operator. The A operator answered the subscribers request. After the subscriber gave the number "Main 1234", the A operator would plug into the Main exchange where a "B" operator would take the call. The A girl passed the "1234" to the B girl, who made the final connection. Because in big cities most calls were out of the exchange all calls were handled this way. Early dial automation continued this pattern of A and B handling.
Reply to
Lisa or Jeff
Loading thread data ...

I know that there's anecdotal evidence that suggests that some panel offices did indeed have over 10,000 lines. However (comma) ...

Wow {dusting off very rusty memory cells}, I actually knew this CO back in the years when I was more of a phone 'enthusiast'. ;-)

I'm trying to think back, and IIRC this (465) was panel in the early 70s, which would imply that it was converted from manual to dial (panel) prior to WWII. My impression of the age of the building was that it was definitely pre-war. LOL, I could even drive you to where it was at the time. ;-) It could have been a #5 crossbar, making it possible that it was still manual in

1947. I do know that Chicago had some manual service into the mid 1950s, as did Omaha, home of Ma Bell's first full-scale panel deployment.

See my final conclusion below ...

I was far more familiar with the Floral Park office, not too far east of the Hollis office, which was at the time a #5 crossbar, to which I've referred here in the Digest on occasion. This was unique in that it served both 212 and 516 prefixes at the time. But back to the subject ...

(Now I'm REALLY dusting off the memory cells, remembering the architecture of panel ...) While it could be argued that simply adding a 21st. final frame to a panel office would expand the capacity to 10,500 lines, I'm having a difficult time figuring out how it could be shoehorned into the set-up of the incoming frame, and addressed by the revertive pulsing, which was very much hard-coded into the selection of one in 10,000 lines.

There were two revertive-pulse operations associated with the incoming frame, which selected the final frame. The first was a one-in-five selection which selected the brush to be used or 'tripped', and the second was a one-in-four group selection which selected one out of 20 final frames for a full office. IIRC that second one-in-four was via a commutator which had rather distant spacing. If you listened to the cadence of the pulses coming back from a distant panel office, the very first was rapid, and the second was quite a bit slower. A vacant connection to the final frame in question was then selected within that group.

In order to add a 21st. final frame, that group selection would need to be a one-in-five, which implies that a fifth group would need to be added to all banks of the incoming frame with a modified commutator and quite a bit of wastage in the process.

I think maybe what happened is more along this line, which is purely speculation on my part since I wasn't even aware of any

8 digit local numbers until long after I knew how a panel office worked.

My guess is that new 'machine switched' offices were installed with 10,000 lines and converted manual offices were truncated or renumbered to 10,000 lines. However, I think that some of the legacy 8 digit numbers may have been dialable via the Panel Call Indicator on the B board of the legacy manual office.

I don't remember many of the details of PCI, but I do know that it was quite versitile and used for things other than just an indication to a B operator. I'm sure it would be possible for a register-sender in an early panel office to be provisioned to accept an eighth digit, with timeout in ambiguity cases, using simple relay logic.

My guess is that when manual offices were converted to that newfangled machine switching, the numbers above 9999 were retired.

Your citation of 1947 listings of 8 digit numbers does imply that they were dialable, since there was a LOT of panel in service in those days and the post-war deployment of the #1 crossbar was replacing the remaining manual offices. I doubt very seriously that a dial subscriber had to go out of procedure and dial 0 Operator to be connected to those few

8 digit numbers. ;-)

My final conclusion is that HO5 (465 or legacy HOLlis) was most likely still manual in 1947, soon to be converted to crossbar, and that during the phase-in of dial, but prior to DDD, yes, some 8 digit local numbers were indeed dialable.

Anyone care to refute (refudiate?) this ?? ;-)

Reply to

I understand that two large suburban exchanges in the Phila Metro area didn't go dial until circa 1962 (FLanders in Upper Darby and Willow Grove). What puzzles me is that FLanders was apparently dialable using call indicators, but Willow Grove was not dialable and required an operator to connect the call. A 1959 Phila White Pages did not have any dialing instructions, to my surprise.

Converting a manual to dial office was extremely expensive and time consuming. It took a number of years just to convert Manhattan, and then longer for the outer boroughs. Further, some outer boroughs weren't even developed until the postwar era.

The postwar era saw tremendous growth in the demand for phone service. The Korean war and Cold War diverted telephone equipment. I saw a postwar photo of a suburban Phila exchange (Ambler, PA) where additional switchboard positions were clearly shoe-horned in a building--one position was squeezed at an angle in a corner as part of the multiple. The neighborhood served by the HOllis exchange had postwar growth, perhaps they had to squeeze in eight digit numbers as a temporary expedient.

I've heard the New York Public Library has old phone books going back years. If they have one for Queens it may explain this.

The Bell Labs history notes that a new panel exchange was installed in the NYC area around 1950. This is surprising. Presumably they wanted compatibility with other equipment rather than using a No. 1 crossbar. Also they said new pieces of panel were added for many years.

In Phila, there were two city offices that served the nearby adjacent suburbs (ESsex in Cheltenham was served out a lower NE Phila office, and MElrose in Melrose Park was served out of the WAVerly office). What was curious is that calls to and from those suburban exchanges were billed at the suburban message unit rate even though they came out of a city office. That is, someone in downtown Phila calling the lower NE would be charged as a city call, but calling Cheltenham meant message units. I think those arrangements continue to this day, though they call it "Measured Service" and they've reduced the charges from the past.

I'm starting to wonder if the HOllis-5 office was manual. Did panel switchgear store the entire phone number before acting on it, or did it translate the exchange and then route the rest of the dial pulses to the desired exchange? I understand your point of the frames, but perhaps only the incoming storage register had to be larger.

The B operator's display is shown in many articles about panel (though I don't know where there's one on the web). The display actually has capacity for _nine_ digits: the leading 0/1 for eight digit numbers, the four numbers, and a party line suffix letter. (I suppose these could've been dialed, too.)

That the display has this capacity is what makes me believe eight digits were dialable.

Agreed. In the postwar era, while many subscribers had little change to their numbers, others did. Sometimes it would get tricky--I've seen special announcements printed for a community detailing the new phone numbers.

Reply to
Lisa or Jeff

It acted as soon as it had enough information to act. This was very innovative for the technology of the day.

IIRC, the action on the dialed digits was somewhat overlapping on panel. It was definitely not direct control by any means. If you would dial slowly when calling one panel office from another, you could often make out the sounds of the progress through the switch train. The cadence of the sounds was quite different when calling a #1 or #5 crossbar from a panel office than it was when calling another panel office, and this made it somewhat easy to tell which offices were panel and which were crossbar.

Of course if a tandem office was involved, things got interesting.

But anyway ... lots of this is from memory so some of the details may be a bit off ...

Nothing really happened until the three digits of the office code were dialed, then the register-sender called upon the translator or decoder to get the orders on where to set the district (and office frame, if used) frame to get to the terminating CO. You could usually hear some scratches and clunks as this happened.

The rest of the process was well known by 'enthusiasts' as the revertive pulsing was 'hard coded' to the particular directory numbers. The register-sender in the originating office always 'knew' exactly where to set the incoming and final frames for a particular 4-digit number. To hear this you would pause after dialing the office code.

When you dialed the fourth digit, the first digit of the line number in the distant CO, you would hear a 'scratch' of revertive pulsing as the brush of the incoming frame was selected. Then when you dialed the fifth digit, you would get more of a 'clack-clack-scratch' as the incoming frame advanced and the final frame brush was selected. The last two digits gave you that 'scratch-scratch-CLUNK' as the final frame advanced and connected. IIRC this was two separate operations, the second at a lower speed.

#include ;-)

Party lines were all but gone in the city when I became interested in such things, but I do remember references to manual exchanges using lettered suffixes instead of assigning party line subs their own directory number. This would mean that a full manual office could very well handle well over 10,500 subscribers and that yes, there could be numbers of the pattern FNOrd 2368j and and even FNOrd 10015j and such. Now could those numbers be dialed by the early dial subscribers? I have no clue! ;-) 'Maybe' is my best guess. ;-)

PCI was very powerful, and could handle arbitrary digits and such as long as the equipment was able to handle it. When dialing from a panel office to crossbar, it was very easy to confuse the sound of revertive pulsing (the cadence of which was far more consistent in crossbar offices) with that of PCI.

Revertive pulsing could handle that 1 of 10000 selection, and in cases of panel offices, could not do such things as deliver an office code. (Yes, purists, I do know what 'high-5 is'.) ;-) PCI was what was typically used when routing calls from a panel office via a tandem office. The XBT crossbar tandem (think of a #1 crossbar on steroids) typically received calls using PCI and then sent them to the terminating CO via revertive pulsing. There was also a panel tandem, and I'm sure it worked more or less the same way, but these were all gone by the time I got interested in these things.

ObTrivia: The only practical way of subscriber dialing between panel and step offices was to go via a tandem which spoke all of PCI, dial pulsing, and revertive pulsing.

My guess here is that your Willow Grove office used some kind of manual equipment (AE?) which did not support PCI or else was planned for conversion and never intended to be dialable. I would, however, think that it would be cost-effective in any case of medium to long-term continuation of manual service to convert to PCI, since the labor saving would be significant.

It's been my impression that such offices as Floral Park and Bayside (coincidental pun, located right by Bell Blvd.) ;-) were postwar expansions intended to handle just this. I would put money on the fact that some of the territory now served by the Floral Park office (physically outside of the city but serving a number of 212 numbers in my day) were originally handled out of the Hollis office.

My hunch here is that they wanted to use up existing inventory of panel gear. Remember that the #1 crossbar was essentially a drop-in work-alike of a panel office using newer technology.

I do know that many panel offices were built out to the full

10,000 lines using panel technology instead of being upgraded to crossbar.

I also vaguely remember a case in lower Manhattan (Broad St CO, maybe?) where a to-be-upgraded panel switch was left in service and the 1E intended replacement was put in service to increase capacity.

***** Moderator's Note *****

PCI is "Panel Call Indicator", the system that lit the indicating lights on the "B" operator's board.

Bill Horne Moderator

Reply to

When I joined Bell Labs in 1959, I had a series of classes about introduction to the Bell System. In one of them, the instructor said that Bell Labs invented the panel switch because they wanted the flexibility of an X-Y switch like crossbar, but did not want to pay royalties to the inventor of crossbar. He said that panel was clunky and more prone to breakdown, but it beat paying royalties. I don't know who was the invetor of crossbar.


Reply to

The story I heard and believed was that Ma Bell did not embrace STEP technology because they did not want to pay royalties to Strowger and company. Add to that the fact that direct-control step does not scale well, of course. I never heard that story about crossbar. Of course I never had the opportunity to attend an orientation session at Bell Labs. ;-)

I guess they did eventually employ SxS for many smaller communities in their service areas, and even some mid-size cities as well, with Des Moines being an example.

I always assumed it was Bell Labs !! ;-) At least in the context of telephone switching that is.

I don't think anybody is gonna argue with you over that. ;-)

You must admit, however, that for its time (the 19-teens) it was very innovative and high-tech, even though it's been said that the design was inspired by Rube Goldberg. ;-)

The story (legend) that was circulated among 'hobbyists' in long-ago decades was that the panel switch was originally an automated B board and was designed as an electromechanical representation of a 'panel' of jacks with the selector rods and brushes representing the operator's arm and plug. The revertive pulsing was representative of the A operator telling the B operator where to plug in the cable. ;-) 'Keep going, keep going, no, higher .. keep going .. STOP!'

IIRC there were a few of these B board panel installations in the 19-teens, long before the first full-scale panel roll-outs in Omaha and Paterson.

Reply to

The Bell area of Los Angeles was all SxS in until after the war.

Reply to
Sam Spade

Beginning in 1965, I lived in Salem, NH and worked in the Lawrence, MA area. Both Salem and Lawrence were NE Telephone and both had step offices.

Both cwntral offices allowed last-5-digit dialing when calling phone in the same respective city. Both 5- and 7-digit dialing worked.

In Salem, the office codes were 893 and 898. So a call to 898-8123 could be dialed as 88123. Then, the NH PUC ordered that our local area include Nashua, NH, with office codes begining with 88 (882, 883, ... 888, 889). To make this work, NE Tel had to change the Salem numbers of the form 898-8xxx to 893-0xxx, so that any dialing beginning with 88 would be directed to Nashua. 5-digit dialing still worked from Salem to Salem.

In Lawrence, the office codes began with 68 (681, 682, etc.) 5-digit dialing existed for a long time. Some long time residents were so used to the 5-digits that they would give their number as "68 (pause) 23456" instead of "682 (pause) 3456".


Reply to

I thought panel offices used revertive pulsing, i.e. after the full number was dialed the distance office started pulsing back to the originating office until the originating office said "stop" and went on to the next digit similarly. The first No. 1 crossbar offices were arranged to simulate panel offices in their signalling, leading to the apparent absurity of two No. 1 crossbar offices each simulating panel's revertive pulsing. The first No. 1 crossbar was intended for use in an all-panel environment.

Wes Leatherock snipped-for-privacy@yahoo.com snipped-for-privacy@aol.com

Reply to
Wes Leatherock


How about Houston, Dallas, Los Angeles, Oklahoma City, Fort Worth, Tulsa, Wichita, Little Rock.

Wes Leatherock snipped-for-privacy@yahoo.com snipped-for-privacy@aol.com

Reply to
Wes Leatherock

As I stated about Los Angeles several messages back. ;-)

Reply to
Sam Spade

The Bell Labs history says that by 1900 manual switchboards were highly developed and from the subscriber's point of view, easy to use. In contrast, the Strowger switch at that time wasn't as developed, such as requiring subscribers to ring manually. It was also expensive since it was a large capital outlay. Manual offices didn't need operators--saving wages--during off peak hours. Only very tiny offices were seen as candidates for dial since there wasn't enough traffic to justify even a single operator. (However, for many years Bell had such light duty operators in private homes, see the Mountain Bell history for a description.)

The Bell Labs history says Bell made many improvements in the Strowger switch for Bell Service.

Note that by the time dial was perfected it took about 50 years to convert all Bell System lines to dial (Santa Catalina Island. being the last). That's a long time. Conversion to dial was a major undertaking. Some manual exchanges were converted directly to ESS.

Step by step handled the majority of dial lines of the Bell System. Plenty of mid sized cities had it. SxS lines peaked in 1974.

I wonder if its developers used tools like Boolean algebra to work out the logic.

I believe also in Newark, NJ. This made sense, as it gave Bell engineers experience before the big rollout in NYC.

  • * * *

Five digit dialing lasted for a long time in small towns. Yes, residents did refer to their numbers in the above fashion. Also, stationery listed a business phone as (311) 55 5-2368, with a space in the exchange.

A resident of a such a small town told me the phone company told people to use 7 digits and stopped referring to 5 digits. But 5 digits continued to work until the exchange was cutover to ESS. (He hoped to get a switch unit after the cutover but the old gear was kept quite secure. I would think its value would only be scrap by that point in time.)

Reply to
Lisa or Jeff

I was never in an actual C.O. in Des Moines, but in the mid 60's, I did belong to an Explorer Scout troop that met at the Bell training facility downtown.

There was a _small_ practice switch where we met -- the switching elements did -not- resemble any of the pictures I find on-line of a Strowger SxS relay, was definitely -not- crossbar, and I'm not having any luck finding a decent picture of either a 'rotary' or 'panel' switching element, to validate my recollections against. Recollection is a gizmo that processed a two-digit chunk of a number -- a disk element that rose on a rod for the first digit, and rotated for the second one. *guessing* (=very= roughly, at a 45-year remove) at a circa 1-1/2" dia, and a vertical travel of 6-8".

Reply to
Robert Bonomi

Or Carbondale IL, Ann Arbor MI, or Centerville IA?

formatting link
Neal McLain

Reply to
Neal McLain

It's always been my understanding, and I'm fairly certain of this, that on the panel offices which were common in the late 60s and early 70s, the dialing and outpulsing could be overlapping, in that once the office code was dialed, the register-sender of the originating panel office seized a trunk to the terminating office and as digits were dialed, the call progressed accordingly.

The revertive pulsing corresponded to the four digits in that each four-digit line number had a definite sequence of revertive pulsing, but it was not four actions, not one per digit, it was five.

A true panel office was very noisy in that the sounds of the office could be heard during and immediately after dialing. I really think that a lot of people who are otherwise very CO savvy are forgetting this. Panel offices were also notoriously noisy during the connection. Contacts were often microphonic. One sound was kind of a cross between a 'clang' and the sounds of chains rattling.

You could clearly hear the progress of the call when you were dialing from a true panel office. Not as much from a #1 or #5 crossbar. Most users simply dialed away, but those who were interested in such things would often pause and listen closely when dialing.

Many phone 'enthusiasts' were quite familiar with the sounds of the quirky panel offices, and knew the following operations quite well:

  1. When the thousands digit was dialed, there would be one burst of revertive pulsing which could usually be heard. It was heard as a brief 'scratch', and was that one-in-five selection of the brush on the incoming frame.
  2. When the hundreds digit was dialed, there would be TWO bursts of revertive pulsing, the first being that one-in-four group selection as the incoming frame selector advanced, and the second as a one-in-five selection of the brush on the final frame. This would sound like a 'clack-clack-clack-scratch' when calling from a panel office to a panel office, and more like a 'scratch-scratch' when calling from a panel to a crossbar office.
  3. The last two digits, of course, selected the one-in-100 position of the final frame selector. IIRC, the second 'scratch' in this sequence was a bit slower when calling a panel office than when calling a crossbar office.

IIRC, the register-sender (or what it's properly called on a #1 crossbar) did record all digits before seizing the trunk to the distant CO, but I'm not 100% sure of this. There may very well have been some overlap, as I remember that when the 1E offices started coming out around 1970, a call from a #1 crossbar office to a crossbar/panel office went through very quickly, where a similar call from a 1E office had a delay of about a second after all digits were dialed.

I don't know about any of these except Los Angeles, and it's been my understanding that quite a bit of the El-Lay area was indie using AE step, and Ma Bell simply went along. ;-) I do know that there was quite a bit of #5 crossbar in Orange County in the early 1970s, as well as some 1E making inroads into the area as well. I vaguely recall much of Santa Ana being step as well.

Our own Pat has eluded to the Wabash office in Chicago being step, but I kind of question this, since I know a lot of Chicago was panel and #1 crossbar from the 30s through the

Reply to

At ACTA and TCI phone shows, sometimes collectors bring in working switches to demonstrate, including panel. Visitors can make calls on them. Sometimes different exhibitors link their switches together. (Sometimes they use electronic black boxes to accomoplish the interface).

When I watched the panel frame work, the motor and rods were quiet, certainly quieter than the ratchet of the step-by-step switch or the relaxy clacking in a cross bar. Of course that was but a single frame, maybe a room full of them and associated might be different.

Check ACTA and TCI websites for shows--they are held all over the country and Canada. Worth a visit if not too far away.

I think LA was step because it had many independents that Bell took over and in that particular case it was easier to leave it as is. But I suspect given the size and growth of LA it would've been panel or #1 crossbar to facilitate inter-office networking and trunk utilization.

They attempted some SxS front end common control in LA to improve efficiency. SxS also had something called the "graded multiple" which improved trunk utilization.

Downtown Phila had some SxS units serving certain Centrex users. Oddly, SxS but not panel could service Centrex. On a small organization that had centrex via SxS, on interoffice calls the first digit was absorbed and users need only dial the last three digits (this was not well known). The organization's switchboard was a traditional cord unit, too; most Centrex's had consoles for the operators.

Reply to
Lisa or Jeff

Many were converted to 5XB much sooner than ESS became viable. That also prevented five-digit dialing.

Wes Leatherock snipped-for-privacy@yahoo.com snipped-for-privacy@aol.com

Reply to
Wes Leatherock

Alas, there used to be a web site with recordings of all those switch sounds, including panel.

Reply to
Sam Spade

In the early 1970s, Santa Maria CA had a step switch (General Telephone at the time). They had 5 digit dialing. I worked in radio stations in both Santa Maria and San Luis Obispo (Pacific Telephone). SLO had 7 digit dialing. Santa Maria is right on the county line. The radio stations in Santa Maria I worked for would give business telephone numbers as 5 digits. Listeners across the county line could not dial the numbers that way. Santa Maria's switch would accept 7 digit numbers. I always thought they should broadcast 7 digit numbers to be useful to the largest number of listeners. Of course, when 7 digit numbers were broadcast, the prefix was WAlnut 5.

Also at that time, you had to go through an operator to call from SLO to Guadalupe, a town west of Santa Maria. I don't know if Santa Maria could dial Guadalupe direct or if, perhaps, Guadalupe still had a manual switch at that time.


Reply to
Harold Hallikainen

{ ... ]

The first dial office in Oklahoma City, installed by the Bell company in 1920, was AE step equipment ordered through WE as required, since WE did not make step equipment at that time.

In 1927, the rest of the city was coverted to dial with WE SxS equipment.

The AE equipment remained compatible with additions to the first (AE) office being made with WE equipment. In the late 1950s I remember we put out a news story that the ringing sound would be different in that office because the old (AE) ringing machine was being replaced. It was being replaced, not from failure, but because with the continued growth of the office it had reached its capacity. Ten or 20 years later the entire office was replaced with ESS.

It was like a picture out of history books to see the AE line switches in their glass cases. Of course all the later additions used line finders.

Wes Leatherock snipped-for-privacy@yahoo.com snipped-for-privacy@aol.com

Reply to
Wes Leatherock

Cabling-Design.com Forums website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.