How did Harley serial number their bikes?

This is the number system used by Harley-Davidson circa 1916 into the 1960s.

As a little background info, every year from around 1916 onwards Harley-Davidson made up a special reference model "Layout" sheet. For example the 1925 sheet is titled: "Layout of 1925 Models & Model Numbers." It is very complex and includes: "Sales Dept. Model Nos., Factory Model Nos., Motor Equipped, Motor No., Motor Assem. No. & Frame Number." Later sheets were less complex. Since these were intended for internal use they contain inside info not generally known. Good stuff indeed!

The 1925 Layout Sheet clearly states near the bottom of the page and here I quote: "Note: Motor & Frame numbers start with #1000 and continue consecutively regardless of models...."

It also states this little factoid which I believe refers to the belly/line bore number: "(CRANKCASES NO. 25-1000 & up)."

Clearly by Harley's own Layout sheet engine serial numbers and crankcase numbers in the old days started with #1000 (not 1001).

This numbering method continued throughout the 1920s, 30s, 40s, and into the 1950s. For example the 1938 Layout sheet reads: "motor numbers start with 1000...." and the 1950 sheet reads: "motor numbers start with 1000...."

However, when we reach the 1957 Layout sheet there is an abrupt change in the numbering system. For the 1957 model year it states: "Motor numbers start with 1001..."

No reason for this change was given.

I should point out that both the 1000 start and 1001 start were for normal production models: JD, VL, EL, UL, FLH, XL, etc. Racing engines had different start numbers: i.e. "500," "2001," etc.

In 1960 the serial number system became more complicated swapping starting numbers between 2001 for even (1960) and 1001 odd (1961) years. I do believe that Bruce Palmer mentions this in his restoration book.

In 1970 Harley dropped the old time method and went to the govt. VIN system which has no coolness to it at all.

So you can see this wasn't a trick question, but it has a rather tricky answer. So, if you run across or see a "1000" number don't assume it's FAKE or a Factory only special. It could be authentic normal pre-1957 production.

The obvious question that comes to mind:
Where are all the yymm1000 issues?

(YY meaning year, and mm meaning model.)

Is there a single surviving example?

How horrible if my ninth one off the line in '49 is actually the tenth!

OK, not horrible, but it doesn't roll off the tonque as well.

Once again the rule is: never say always.

....Cotten

The obvious question that comes to mind:
Where are all the yymm1000 issues?

(YY meaning year, and mm meaning model.)

Is there a single surviving example?

How horrible if my ninth one off the line in '49 is actually the tenth!

OK, not horrible, but it doesn't roll off the tonque as well.

Once again the rule is: never say always.

....Cotten

Exactly. A small point that changes the status of every bike that rolled off Harley's production line over the years.

Plus the FACT that a "1000" bike isn't automatically a fake.

Whether or not there are any 1000 bikes out there I can't say. I don't personally know of any although people on this thread have already mentioned seeing old photos with 1000 serial numbers. I can add that at least one 1000 serial number shows up in old Wis. motorcycle registrations from the 1930s or 40s.

Apparantly this basic truth about Harley's numbering system has been widely misunderstood over the years. That's why I wanted to know first what the common perception is. From the tenor of the posts it seems that if a 1000 number bike showed up at a judged show it may likely have been disqualified as a "fake" since most people seem to think 1001 was the starting number. By Harley's own reckoning 1000 numbers are authentic and not fake.

To back that I have found another source. The handbook "Q&A" by your "Uncle Frank." (Revised Army Edition 1945), p. 177, and again I quote: "All Harley-Davidson motor numbers start off a new season with 1000."

I have always thought that the same sequental number could only be shared between different lines like Big Twins vs Small Twins/Singles. According to the above, there could exist a 38EL2475, a 38U2475 and a 38W2475. But is there anyone out there who owns two or more motorcases like that or at least has seen a matching pair?
Fiskis

If the motorcycle serial number was stamped on the motor when it was built and the motor then put into storage,then pulled at random when that type of motor was required for final assembly.

What would happen with unnumbered spare part motors? Would they have a special place in the stores? and what would happen when the new year started would you get 47 models with 46 engine numbers?

Just Curious Pete,

I suppose sales of complete engines has been small, as they must have cost half the price of the bike. Probably a unstamped motor with 46 belly numbers would, if leftover, be sold as a spare part for '47 models (or older).
Those early modelyear bikes I seen, usually carry leftover parts, so my suggest is that the 47-stamped cases came first and then they dressed up the fortysixes with '47 parts. Like the prototype '47 in Palmers book.

For some reason people hesitate to release their serial numbers, is there a risk that someone picks up an Alabama title on that same # and claim it is their bike?

It would be interesting to check the VL Register if there is a match for your 36 Knuck'. I believe there is a 36EL register somewhere. And there is a DL-club. So after a little research you could own three thirty-sixes with the same sequental number.

See you at the AMCA Europe meet (or bust)

Fiskis

Fiskis: Damn, I have a hard enough time explaining the model designations to new riders as it is. Now I have to think in odd years and groupings for true serial#s? No wonder we have so much trouble with DMV. And that is just for the honest ones. Sharks are every where. Still willing to learn but head is a bit fuzzy now.

What would happen with unnumbered spare part motors? Would they have a special place in the stores? and what would happen when the new year started would you get 47 models with 46 engine numbers?

Peter,
"Spare part" motors were not numbered. It is not unusual to find, say, a 1941 VIN on a motor with '40 bottom numbers. But you would never find a 1940 numbered engine in a machine sold as a 1941. The MoCo was smart enough to know exactly how many 1940 engines they would need before starting the 1941 models.

Chris.
I know that spare part motors were not numbered, I also know that bottom numbers can be dated the year before the Motor number(VIN),I have motors like this.
I asked the questions to stimulate Further discussion on Harley Creations information that motors were stamped with the sequential VIN number when built then used at random to build completed bikes.

Pete

[quote="peter reeves"]Chris.
I know that spare part motors were not numbered, I also know that bottom numbers can be dated the year before the Motor number(VIN),I have motors like this.
I asked the questions to stimulate Further discussion on Harley Creations information that motors were stamped with the sequential VIN number when built then used at random to build completed bikes.

Pete

That is exactly what happened. The same thing happened after line bore number were applied. If the VIN was applied to line bore numbers in a precise manner they would be close together. Let's say a cart held a hundred sets of crankcases. The one stamped 1000 would be the first on the bottom of the cart. 99 others would be on top of it. The cases were removed at random from the top down and engines assembled. So a set of cases from the top of the cart would receive the 1000 VIN. This is evident with my 1004 engine with a line bore number of 1098.

Chris.
I know that bottom numbers and engine numbers did not follow an exact sequence I still find it hard to believe that completed bikes would have left the production line with out of sequence engine numbers. It just doesn’t sound logical.
Harley some times gave engine numbers with regards to spare parts, using the words “with motor number below ”.
If bikes left the line with random numbers they would not have been able to say below this number???

Pete.

Not to sound flip, but we're not talking the logical way of doing things, we're talking about the Harley way of doing things.

The serial numbers of motors put into bikes weren't wildly out of sequence. According to John Nowak the numbers might have been 20 or 30 off sequence as motors normally didn't stay in the storeroom very long. Maybe one day's production once they got going.

I asked him about that photo of stockpiled motors and if numbers could ever be "hundreds" out of sequence and he said, "No, never." That the only time they stockpiled motors like that was before they started motorcycle production and even then they're weren't that many. (I gotta count how many are in that photo).

That material comes out of a exerpt from a interview with John in 36EL Registry March/April 1995, p.6-7. The information is pretty exact. Apparantly bikes normally left the line with out of sequece numbers but not wildly out of sequence. But that still could happen, however, if motors were kept aside and repaired. Seems to me there was a question like that about the early 36EL numbers.

Again: there was no system of putting motors into bikes in a sequential motor number basis. But later the motor/bike cards that accompanied each motorcycle down the line were put into sequence by serial number.

Your point about the spare parts is a good one. Do those "under" a certain motor number references refer to just motor parts or cycle parts too?

Chris.
I know that bottom numbers and engine numbers did not follow an exact sequence I still find it hard to believe that completed bikes would have left the production line with out of sequence engine numbers. It just doesn’t sound logical.
Harley some times gave engine numbers with regards to spare parts, using the words “with motor number below ”.
If bikes left the line with random numbers they would not have been able to say below this number???

Pete.

I have found a shop dope 137 with a numbers below 35RL3466 entry that refers to a 1935 RL transmision.
But it also mentions 15 machines with engine numbers below it with a spread between 3222 and 3466 that already have the latest part fitted.
You Might well be right.
Pete.

If bikes left the line with random numbers they would not have been able to say below this number???

Pete.[/quote]

Peter,
If the parts manual says that a change was made to the timer starting with engine number 1222 that is what it means. The engines were numbered is sequential order. Then placed on the carts. They were removed from the carts in random order. Does this make sense to you now?

Chris.
Read my last post .
I specifically referred to a shop dope relating to the transmission on a 35RL.
Obviously if they had said number before in relation to an engine component such as the timer it would relate to some thing fitted to the motor. But if they said before engine number in relation to the transmission this most relate to completed machines.

Pete.

Pete,
I was not responding to that post. Read the quote.

Other examples similar to Peter's:
A Shop Dope service bulletin issued between September and October 1928 discusses a new pin fitted through one of the breather holes in the large generator bevel gear and the bronze bushing supporting the right side of the generator drive shaft and states: “This pin is fitted in all new motors with numbers above 29D3403” and “This bushing has been fitted in all motors with numbers above 29D3404”.

Shop Dope #112 dated December 17, 1934:
“1935 Twin model motors are fitted with T-slot piston. 1934 Twin models are also fitted with this type piston starting with motors numbered 34-R-4000 and 34-VLD-9000”.

With regards to the discussion about different models from the same year having the same serial number, I find this unlikely. The total production of D series motorcycles in 1929 was 6,856, but I have a motor numbered 29D8152. And the total R series production in 1934 was 1,992, but the shope dope above mentions motors numbered 34R4000.

Looking up the Shop Dope service bulletins is a good idea. I see a couple more listing erratic (throwback?) serial number sequences: No. 142 (36EL) and 153 (37EL).

But lets use Peter's No. 137 (1935) listing "R" models because that refers to completed machines and not just motors.

No. 137 lists 15 machines below 35R3466 with updated transmissions. The motor SN's on these bikes range between 3222 and 3465. That is a deviation of almost 200 motor numbers out of numerical sequence; FAR greater than the 20 or 30 that John Nowak said.

Unless John was wrong and the out-of-sequence deviation of motors going into bikes was much greater than 20 or 30 motor numbers off, the other possible explanation is that these motors were held back for other repairs.

According to John Buchta they often did that too. If when block tested they made gear noise or something they were held back and fixed. But since they already had their motor numbers stamped on them, these slow-pokes could be far behind the pack number-wise by the time they reached the motorcycle assembly line. In this example some 200 numbers behind!

Another possibility is that things changed over the years and early on there was a wider out-of-sequence range of motors going into bikes than later on. John Nowak started in 1936 and didn't have much authority until the 1950s.

They talk about the Da Vinci Code as being hard to figure out. That ain't got NOTHING on Harley's esoteric crack-brain numeralogy.

No wonder Harley called their service bulletins "Shop DOPE." It's enough to rattle anyone's brain!

Johnny,
The published Harley-Davidson figures are in no way correct for year models. They were compiled years ago by a secratary. She came up with those figures by going through the quartely shareholders reports. She took the reports for Jan-Mar, Apr-June, July-Sep, Oct-Dec and totaled up the number built. The only problem is that this is the last half of one model year and the first half of another and not the the total for any one model year.

Johnny,
The published Harley-Davidson figures are in no way correct for year models. They were compiled years ago by a secratary. She came up with those figures by going through the quartely shareholders reports. She took the reports for Jan-Mar, Apr-June, July-Sep, Oct-Dec and totaled up the number built. The only problem is that this is the last half of one model year and the first half of another and not the the total for any one model year.

That's very interesting and something that I had not heard before.

What published source with incorrect production figures are we talking about?