This is Part 2 of the revised Dyer section of Harpguitars.net.  This section will compare the timeline of the Dyer instruments gathered from BMG journal advertisements with clues from surviving instruments, including visual features and label information – all in a continuing effort to come up with a plausible serial number sequence and ultimately date these instruments.  The reader should first be familiar with the serial number sequence as presented by Robert Hartman in the latest “Centennial” edition of his book series.  As Bob freely admits in the book, this sequence is plausible only insofar as it adheres to a consecutive sequence for all instruments in the Dyer line, but seems implausible in light of the advertising provenance (specifically the mandolin family line).

One thing that has frustrated Bob and others researching this subject is the spotty, random sampling of ads that were known from early 1900s journals – specifically the “BMG” magazines (various journals which highlighted the activities of Banjo, Mandolin & Guitar players).  During several months in 2008, I was finally granted access to, and allowed to photograph, a nearly complete run of Cadenza and Crescendo magazines.  I expected a gold mine of early American stringed instrument history, but quickly realized the limitations. These periodicals are unfortunately not the "be all and end all" of research that some of us were hoping for.  Of the many hundreds of stringed instrument makers and companies, only a very small percentage regularly advertised in, or were even mentioned within the pages of these magazines (generally spanning 1890-1920s).  But luckily for us - in fact, by sheer dumb luck - the Dyer Company was one of these.  For many long periods, hardly a month went by in which they were not featuring one of their "harp" instruments in one and usually both of these two important periodicals.  These ads - of which we have had access to only a scant random few until now - have long been referenced by Bob in his Larson Brothers books, and are invaluable in piecing together the history of the world's most popular harp guitar of all time.  Additionally, in early 2009, beyond my wildest dreams, scholar Jeffrey Noonan published a ridiculously detailed and meticulous Bibliography of not only the full Cadenza and Crescendo runs, but of virtually every periodical he had access to for his seminal work The Guitar in America of 2008.  This bibliography contains hundreds of harp guitar entries - the trick now is to gain access to the rarer (as in perhaps one copy known) periodicals.  Because of these recent findings, we have had to re-think the old Dyer serial number system.  I also took this as an opportunity to re-examine all of what we know about the Knutsen and Dyer instruments.  The results surprised me.  I have not only completely re-vamped the Dyer dating scheme that Bob has worked so long and hard on, but created four different hypothetical timelines, any one of which may be closer to or further from the truth than the others.  Each one is an attempt to better fit with both the advertisement evidence and the many evidential anomalies.

I am certain that there were different serial number series for the harp guitars, harp mandolins, mandolas and mandocellos - in three or four separate sequences, not one.  The exact years of the specific numbers are still fuzzy, and my harp guitar dates are just alternate theories.  I don't believe I have necessarily disproven the original Hartman system (at least for the harp guitars), so much as come up with compelling alternatives that fit more of the evidence (though some may seem a bit theoretical and controversial).

But if only hypothetical theories, why mess with the dating at all?  For one thing, it occurred to me that our current dating system is itself full of hypothetical assumptions.  Secondly, as a logistical exercise, I was looking for a way to fit Stephen Bennett’s #669 Style 4 harp guitar into our system – approaching it from a working hypothesis that the handwritten date of “1909” on his label = legitimate evidence (full analysis here).  In doing so, other factors came into play, such as the introduction of the harp mandolin (rife with its own mysteries!) and the very recent (and exciting) finds of Type 3 harp guitars with serial numbers.  These clues – along with plenty of “red herrings” – have kept my head spinning and my dating theories circling each other like a dog chasing its own tail.  I expect the reader will be similarly as frustrated as I as they wade through all the information below.  However, you may find some intriguing new theories and answers when done.

I have found no simple way yet to tackle this.  In fact, I can find no way to legitimately present just one timeline (and corresponding serial number system).  I have narrowed it down to four.  Each column follows an alternate theory. Where they merge, it means that they (and we) agree, as there are known facts or consensus for us to do so.  It is perhaps easiest to follow one Timeline all the way down at a time, and later compare the adjacent options where they diverge.  Following the Timeline chart is exhaustive discussion and analysis of the evidence, clues and guesswork.  If you’re serious about exploring this, you will definitely need to examine the Members Only Dyers in BMG section.

For the purposes of discussion the shorthand term "Dyer" will be used throughout these pages, and should be taken to mean either the company "W. J. Dyer & Bro." (which was William J. and one brother and later three brothers), or more likely, some unnamed company representative.  I will also refer to Larson author and expert Robert Hartman as "Bob" (normally "Hartman" would be used, but we're too friendly for that).

As always, the books - both the Noe/Most Knutsen book and the assorted Larson Brothers books should be considered a prerequisite before reading these articles.  I will update and clarify certain things while bringing my own perspective to the subject but I don't duplicate too much of the book material here.

OK, now let’s take a hard look at the Pros and Cons of these timelines.

First, a recap of the key evidence:

Most is presented on Harpguitars.net, and includes the Knutsen Patents, the Livermore and Gaskin harp mandolin patents, the Knutsen/Dyer flyer, the complete BMG advertising run (all of which can be studied on this Members Only page), Robert Hartman’s serial number list, which we will continue to update below, and the many instruments themselves, including the critical label information.  Unfortunately, some of the latter data may be faulty due to misread, mis-communicated, or poorly photographed or inspected labels.

Before attempting an understanding of the material below, I would highly recommend reading the Knutsen and Larson books, re-familiarizing yourself with the patents, continuing with the overview of models, and then studying the complete BMG advertising run (all of which is now collected on this Members Only page).

Next, a list of conflicting or unproven information and evidence that warrants further study:

1. The Cadenza ad of December, 1901 showing a Type 1 Dyer Symphony harp guitar
2. The various harp guitars of Type 2 design that have very low serial numbers
3. The mention of a “1906 model” referring to the Type 2 design introduced in December, 1904
4. The handwritten “1909” of Stephen Bennett’s harp guitar with serial number 669
5. The claim of a patent on all post-Knutsen Dyer labels, and Knutsen’s claim of “sole patentee” on his own labels
6. The unknown date of introduction of the “lower bass point” harp guitars built by both Knutsen and the Larsons
7. The use of “Style 3” for the Type 3 design, whereas “Style 4 through 8” are used for ornamentation designations on the Type 2 design, while Style “1” and “2” have never turned up
8. The first harp mandolin ad which shows a different instrument than the one produced
9. The appearance of harp mandolins between the expiration dates of the Livermore and Gaskin patents
10. Patent claims past 1912, and patent claims on all mandolin family labels (both Knutsen and Dyer)
11. Duplicate serial numbers

And now, discussion of the highlights above.

And yet….not being able to fit the serial numbers into a timeline implied by all of the ads, I’ve taken a new, harder look at this particular ad.

Whatever you (or I) might think of the two scenarios above, let’s call this a draw for now.

Knutsen

Dyer

But another problem with this ad is that it ran for twenty-one consecutive months, then after a five month hiatus, returned for another ten months.  That’s a full three years of production for a Dyer model of which there are only four specimens known!  One caveat about the 5-month interruption: they ran a new ad for their line of strings for four of these months, so it is possible, that it wasn't so much a "harp guitar hiatus" as a "new promotion" of a new product (Sterling Strings).  And yet - it seems more than strange that Knutsen's Symphony harp guitars from this same period outnumber the Larson Type 1 Symphony by over 20 to 1, when Knutsen did little or no advertising.  A reasonably logical answer to this dilemma might be to consider my alternative theory of the previous paragraph – that the 1901 image was in fact representing a Knutsen – and that the six month gap (September, 1903 through February, 1904) possibly coincides with a period where Knutsen was “let go” and the Larsons were hired and gearing up to copy the Knutsen model.  Knutsen was indeed experimenting with many different features in 1902 and beyond, and his quality and aesthetics were only getting worse in general – so Dyer would have understandably become fed up by this point.  It does seem strange that Dyer would simply re-use the old Knutsen ad and not roll out a big new campaign, but there is a similar and even more inexplicable lack of any sense of excitement when they unveil the Type 2 just nine months later.  I thus don’t see any of this as a deal-breaker in the clue department.

Update, October 2011: Now that a fourth Type 1 specimen has been found – the first with a legible Style number (“5” – more on this significance below), we finally have some better clues. Specifically, that two of the four known Type 1 serial numbers are now known – 125 and 127 – which we believe equates to the 25th and 27th instrument built. With the lowest confirmed serial number of the Type 2 being 140, this would suggest that somewhere between 27 and 39 Type 1’s were produced. Update, March 2020: A fifth Type 1 specimen - the earliest so far - has the Knutsen-signed label with serial # 120. The Style may again read 5 - in this case, a small "05." It too has the simple appointments of the Type 2 Style 4.

Bottom line: When taking not only the number of surviving specimens into account, but also trying to fit the early Dyer serial numbers into these two opposing scenarios, I find myself consistently favoring a Knutsen-then-Larson scenario for the December, 1901 through November, 1904 “Type 1” Cadenza ad. (Just remember that I could be completely wrong)

Counterpoint: Bob recently pointed out the question of the Maurer Company which August Larson bought with other investors in 1900.  He wonders if this could have been a business venture specifically to be able to handle a new Dyer contract.  I also find this timing intriguing – however I’m not too inclined to weigh in on it, simply because of my stance above (the low numbers of instruments found).  Nevertheless, this newly-formed company certainly could have been a key factor in the success of the Dyer line.  Something Bob has never really explored and shared with us until recently is the fact that the new Maurer Company might very well have had an unknown quantity of additional employees to assist in production. They could easily have employed workers from Maurer’s factory, for example.  It is also not known whether August’s partners (Longworthy and Lewis) were luthiers or just businessmen.  The Larsons were a legendary “two man shop” all right – but the key word is “legendary.”  That lore came from Bob’s family members much later, and they may have known nothing of the early Maurer years.   Here is Bob's latest write-up for us on the Maurer Company:

B) Type 2 Dyer harp guitars with very low serial numbers

No matter what we ever prove or even agree on for the early timeline, it seems clear that the Larsons built the Type 1 harp guitar first, and then created the Type 2, abandoning Knutsen’s design.  That’s not to say that there couldn’t have been some overlap.  So you would think that as we collected serial numbers, the Type 1s would be found with very low numbers, and the Type 2s would generally have higher numbers than the highest Type 1.  If only it were that simple! 

One problem that plagues us is indirect evidence.  There are a lot of entries on Bob’s old list (and even some on my later list) that came from a questionable source.  Even when we know the source, if we don’t see a decent photograph of the label at the very least, then we are just taking their word for it.  We realize that faded ink or pencil and/or sloppy handwriting can easily cause a number to be misread, and even if legible, someone’s hasty scribble and subsequent letter or email to us of their provenance can yield errors.

C) The November, 1906 ad that declares “1906 model” of the Type 2 design introduced in 1904

Then in November, 1906 – nearly two years later – Dyer announces the “better than ever” “1906 model.”  Yet it features just a poor resolution copy of the exact same woodcut from 1904 (along with a total of five testimonials).  Again, this could just be advertising hyperbole (“let’s pretend it’s a new, updated instrument!”), or it could have some critical, as-yet-unknown significance to our research.  I honestly can’t think of anything – the sixth sub-bass string wouldn’t show up for another couple years, and we’ve seen no significant Larson design features from the early to the later instruments.  Perhaps that is where we should be looking – a switch in bracing patterns or quality. (Note: multiple-specimen Dyer repairmen like Kerry Char have noticed differences in braces: specifically, he has seen the X braces in maple rather than spruce – but we think that this feature coincides with the more expensive Style 7 and 8 and not necessarily dependent on year built. Certainly, more data is needed). Whatever the timeline of the switch from Knutsen-signed labels to non-Knutsen (coming up below), I wouldn’t think this would be simply referring to a new Dyer label and any perceived quality improvement by the public from this technical change.

Bottom line: I originally thought this was a tantalizing and unanswerable clue.  Now, with the "1907 model" example on top of the original "1906 model" example, I would wager that it actually has no significance.  We can put the conundrum behind us.

We get a lot of “provenance” from unverifiable sources.  Often it is a relative who swears that a handed-down instrument was purchased in some specific year – which usually turns out to be unlikely or impossible.  Other times there are written notes from a previous owner – either a scrap of paper inside the case or something added to the label.  A simple rule is to consider this type of information as a possible clue, weighing both its feasibility and integrity, but to never accept any of it as “evidence.”  Such is the case here.  On Dyer harp guitar #669 – currently circa dated 1914 in Bob’s system – there is a handwritten “1909” on the label.  As I discuss in an article on this instrument, it may or may not have been purchased new, but with all things considered, has a good 50/50 chance of being true.  While it is certainly not “proof” I do think it is worth considering as a compelling clue.

For it to be true, we would need to be able to come up with a completely new dating and serial number system that reasonably places #669 in 1909, while fitting what facts we have.  We would also first need to determine if there was logically a way to abandon the current system, which places the theoretical #601 after February, 1912, when Knutsen’s 14-year 1898 design patent expired.  This obvious deduced scenario has been a “fact” during the entire duration of our decades of research into Knutsen and the Larsons – adhered to by Bob Hartman, Noe and Most, and myself.  But we don’t have all the evidence, just the reasonable logic of the theory.  We do have the findings of Knutsen co-author Tom Noe from his exhaustive search at the U.S. Patent and Trademark Office, which may (or may not) have specific bearing on the case – they “prove” some things but not others.  Regardless, I was increasingly frustrated by all the Knutsen and Dyer provenance that didn’t seem to fit the existing 1912 theory.

So, playing “devil’s advocate,” I first asked myself if there was a way Dyer could have obtained the rights to, or ownership of, Knutsen’s 1898 design patent before it expired in 1912.  Unfortunately, the bottom line is, yes, this could have occurred, but according to Tom Noe, did not occur – all of which I will explain next.

I posed specific questions to both Tom Noe and Dyer owner and law professor John Thomas.  Their answers were very illuminating.

Q: Could a 14-year design patent be “renewed” or extended?

TN: Patent terms are not extendible or renewable.  35 USC 154.  Trademarks and copyrights can be renewed, but patents cannot be. When a design patent application is filed, the patentee selects a term of 3-1/2, 7, or 14 years, and pays an application fee appropriate to the term selected.  Note on the two Knutsen patents that they state "Term of patent 14 years." 

Q: Could two individuals or businesses share this patent?

JT: Yes. The patent holder can license others to use it.  (so “share” only in the sense that both are allowed to market an item under the contract – but one is a patentee, the other only a licensee – GM)

TN: (Since) a licensee does rely on the patent owner to enforce the patent if an infringement occurs, it would be reasonable for Dyer to claim that the guitar was protected by a patent. 

Q:  Would there be a record of the Dyer license?

TN: There is no requirement to record licenses.  They are basically contracts.

Q: Would a patent holder continue to use old labels on instruments to imply the old patent is still valid well past the legal period?

JT: Yes. Beyond the Knutsen and Dyer examples, there have been others.

TN: There is no legal requirement to remove "patented" or patent numbers once the patent expires (this is why we see "C. Knutsen's Patents" on the New Hawaiian Family label long after the expiration of the only patents he had).  (This) raises the question of why isn't a patent owner committing fraud by not taking "Patented" or a patent number off the product.  The answer to that is: #1) saying the product is patented is true (falsity is an element of fraud), and #2) it would place an undue burden on a patent owner to recall every product with a patent number on it. 

Q: What about the licensee (Dyer)?

TN: A licensee merely has a right to use (infringe) a patent.  When the patent expires, anyone can use the patent subject matter.  I suppose that Dyer could still say that the design was patented, but couldn't rely on Knutsen for enforcement purposes. 

Q: What does “sole factors” mean on the new September, 1908 Dyer ad?

JT: In this context, I'm confident that "factor" means agent. 
From www.dictionary.com:

5. a person who acts or transacts business for another; an agent.
6. an agent entrusted with the possession of goods to be sold in the agent's.

It's not a use that one hears these days, but it crops up in old business law cases.  Dyer & Bro. were advertising that they were the exclusive agents for selling (this) harp guitar.  They used the plural because the company was "W.J. Dyer & Bro."

TN: Yes, “sole factors” means they were the only agents selling Dyer guitars.

And the $1,000 question…..

Q: Could a (design) patent be transferred back in 1906?

JT: Yes. From the enactment of the first patent act in 1790, patents have been considered intellectual property that can be transferred, licensed, etc.

TN: (Yes, but) the law requires that any transfer of patent rights be recorded in the PTO, both in ownership records and on the file jacket.  The transfer of ownership records don't exist.  So Dyer was a (only a) licensee until 1912…since Knutsen's patents expired by 1912 (14 years from 1898 on the second Knutsen design patent), there were no longer any rights to own. 

I pressed Tom on how certain he was, as my whole “theory” may well have rested on this.  He provided further details of his search and his qualifications for same, and indeed, the case seems to be closed.

TN: A transfer of ownership of patent rights is known as an "assignment."  The PTO maintains assignment records - drawers and drawers of them because most patents were assigned to companies by employee-inventors who had to as a condition of employment.  Even a sale of a patent from one company to another is an assignment.  If a company acquires another company, all the patents in the acquired company's portfolio have to be assigned because failure to assign with a period of time results in an unenforceable patent.   If you have a patent number, you can go to the assignment records and find out who the owner is.  I did that, so I know. 

In the last five years of my corporate practice, I was Intellectual Property Counsel for Danaher Corporation.  Among other duties, I was a member of Danaher's acquisition due diligence team evaluating patent and trademark portfolios of target companies to ensure that rights hadn't expired and that there were no defective assignments.  I only mention this because I spent hours going through assignment records and was quite proficient at it. 

Without an assignment, it is fraud for another, including a licensee, to claim ownership rights in a patent because it is a false claim with intent to deceive the public. 

So there you have it.  It appears that Dyer did not take over the Knutsen patent prior to its expiration in 1912.

Tom has blown my entire theory…or has he…?

Not at all.  I realized that I was exploring a “patent transfer theory” not just in hopes of moving up the 600-series labels (and all Dyer dates) but at the same time to explain why the “Patented in U.S. & Canada ” remained on Dyer instrument labels so far beyond the 1912 expiration date.  (My “patent extension” question was for this same reason).  Since it felt like a related clue I was focused (and stuck) on this theory.  In any event, based on Tom’s answers above, the latter question remains unanswerable with specifics, but perhaps simple overall – Dyer was in a gray area and chose to just continue using pre-1912 labels that included “Patented.”  The law allows the patentee to do so, a licensee might very well gamble on the same “scare tactics.”  As there is no other rational or legal explanation, and this did take place, what other answer is there?

But wait a minute, you’re saying – Dyer didn’t continue a label in 1912, they switched! (from the Knutsen-signed label to the standard Dyer label)

Ah hah!  Wrong.  They had already switched labels much earlier.  Assuming this is true (remember that this is only a theory), wouldn’t it better fit with Tom’s explanation of patent law above for Dyer to have continued with a pre-existing label (burden of recalling labeled products, etc.)?  As opposed to fraudulently creating a brand new label in 1912 after the Knutsen patent expired…which again claimed “patented.”…?  Both options involve a gray area of patent law and marketing, but with the “1912 new label” option appearing so much less tenable, it (to me) bolsters the idea that Dyer had to have switched labels earlier.

But how? 

I can think of many reasons why Dyer would have done this, and how most of the clues better fit this timeline, and even how it all jibes with what we know of Knutsen.  As to how, I simply assumed “Why not? What do we imagine might have prevented such a simple and obvious event?”  I just needed to make sure that it could have happened.

So, a final email to Tom Noe, patent expert, Knutsen expert…

Q: Tom, the final $10,000 question: Is there any reason to think, or insist, (or have legal basis for) that Knutsen had to continue to sign the Dyer labels for the full ten or so years of the license?

I also included for Tom a re-cap of some pertinent facts:

1. Knutsen’s applicable patent was in effect from 1898 to Feb 15, 1912.

2. Dyer licensed this patent from Knutsen starting sometime between 1901 and 1904.

3. Dyer printed up labels, which were sent to Knutsen, who signed them and returned them. These were sent to the Larsons, who entered serial numbers and installed them in the Dyer instruments.

4. All of these labels have serial numbers only in the 100 or 200s.

5.  When Dyer switched to a NEW label - one without Knutsen's signature - the numbers jumped to 600.

6. There is no record of Knutsen transferring his patent to Dyer.

Could not the licensing contract simply have been revised so that Knutsen's signature was not necessary? 

TN: There is no legal requirement that Knutsen had to sign labels, except by contract.  Of course, we have no written contract so we don't know exactly what the arrangement was.  But the fact that the first few labels were notarized tells me that Knutsen didn't trust Dyer initially, and relaxed a bit as time went by and subsequent labels were signed but not notarized.  It is really unusual for a licensor to sign labels or products.  I think this was Knutsen's way of keeping track of royalties, at least at first. 

One possibility is that Dyer negotiated a paid-up license so that the parties didn't have to deal with labels/instruments on a piece-meal basis.  (Regarding the gap between Dyer serial number series) it could be that Knutsen initially signed 500 labels, 300 of which were returned when some new arrangement was reached. So Dyer designed a new label, and started with serial number 601.  That’s one scenario.

Bottom line: your theory holds water.  By the way, paid-up patent licenses are quite common.  What that means is that for some agreed-upon royalty figure, the license is paid up for the term of the patent.  It eliminates all the reporting, paying ongoing royalties, and having the licensor looking over your shoulder.

And so, once again, there you have it.  Understand that we don’t have proof that such a thing occurred.  But also understand that, similarly, we don’t have any proof (and little reason to continue to believe) that the Knutsen-signed label continued all the way up to the expiration of his patent in 1912.

I don’t know how simple this seems to you, or how long it has taken you to read, absorb and ponder.  It has taken me years, months, weeks, and now many dozens of hours to rewrite for the hundredth time – so I know it’s complicated and convoluted, but I hope the reader considers that I’m on to something.

Regarding Tom’s interesting speculation above about the labels jumping in sequence, it might make sense if these “500 labels” were all serialized, but they weren’t – they were left blank for the Larsons to fill in. (Update, Sept, 2011: though now I’m wondering that – since the red pencil/ink of Knutsens signature and the Style and Serial numbers seem to match up, that Knutsen didn’t possibly do all that himself…?)  In any event, I’ve studied American fretted instrument history long enough to conclude that maker’s serial numbering systems are often convoluted and arbitrary.

Now, using the above premise, I’ll re-examine the Knutsen/Dyer timeline.

We know that Knutsen signed the earliest Larson-built Dyer harp guitars, and that the U.S. and Canada patent referred to on the label was his (Knutsen’s c.1899 labels also listed “ England ”).  We can thus safely say that there was a licensing agreement between Dyer and Knutsen in this period, which presumably started when Dyer stopped distributing Knutsen instruments and instead commissioned the Larsons to build a better version.  This must have been in either mid-to-late1901 or in the later 6-month ad hiatus of 1903/1904.

At this point, with the Larsons building substantially more professional instruments, what if Dyer wanted to completely distance himself from Knutsen, whose name was still on the labels?  Might not Dyer’s customer get the impression that Knutsen built their instrument since he signed it?  Dyer would never subsequently credit the deserving Larsons for their work – wouldn’t he have been even more reluctant to have Knutsen’s name on the label?!  Remember that this would have been going on for a full ten-plus years during the height of Dyer’s Cadenza ad campaign. 

As stated above, I find it much more likely that Dyer convinced Knutsen (with financial incentive or otherwise) to re-negotiate their licensing contract.  If such an event occurred, it would not take effect until mid or late 1906 at the earliest, and mid-1908 at the latest, and here is what would have theoretically happened:

1. Knutsen would stop signing Dyer labels and the new Dyer labels would be issued.
2. The “600” series of serial numbers for harp guitars would begin (this almost certainly coincided with the label change).
3. Knutsen would stop using “Symphony” on his labels, while Dyer continued using it (this is a known fact, with an approximate date of this very period. Specifically, circa 1906 he went from “Sole Patentee of the Symphony Harp Guitar with 11 Strings” to “Sole Patentee of the Harp Guitar with 11 Strings.  The theoretical reason being that Knutsen has now licensed the exclusive use of the “Symphony” name to Dyer.
4. Knutsen would still claim “Sole Patentee” on his labels and Dyer would claim “Patented in the U.S. & Canada ” (these are both facts: for Knutsen, the approximate dates of 1906/1907 through 1913/1914 for this label feature are fairly accurate, while the Dyer date is the unknown.
5. Knutsen would begin making only his “lower bass point” (a pointed flare on the lower bass bout) and “double point” harp guitars (this is a fact, with an approximate date of this very period).  Many of these would be shorter ¾ scale instruments.
6. The Larsons would also build a “lower bass point” harp guitar with a short scale.  I had dubbed this a “Type 3” Dyer, and by complete coincidence it was also labeled a “Style 3.”  It was built right at the start of the label change, as every specimen found with a label has a serial number in the very low 600’s (this is a fact, though the date of this label change is unknown).  It is not known who copied who on this design, but it is clear that Knutsen built them from circa 1906/7 to 1913/1914, but not beyond, and that the Dyer version appears to have been only made for a very short time coinciding with the low 600 serial numbers.
7. Dyer would state “Sole Factors” on their ads beginning in September, 1908 (this is a known fact. However, I no longer think it relates to the new license or new exclusivity of the “Symphony” term [though it may].  It more likely means that Dyer was merely the “sole agent” for selling Dyer harp guitars.
8. Dyer would also introduce a harp mandolin (this is a fact, as seen in the September, 1908 ad. However, the pictured instrument – with the configuration of Knutsen’s “double point” harp guitars - is not the instrument that was eventually produced, and the ad ran only three months.).

OK, so are there any logistical problems with any of these eight theoretical events?

No.  Bullet #4 - which incidentally is “Conflicting Evidence” bullet E from our original list far above – is fully covered by Tom Noe’s answers above.  I was always stumped by this issue: How could both Knutsen and Dyer claim that their harp guitar was protected by the patent at the same time – with Knutsen even claiming “sole patentee”?  As Tom explained, Knutsen was still the "sole patentee" – he never transferred it, so remained the sole owner of the patent for its fourteen year duration - and Dyer was still a licensee, allowed to claim patent protection for their licensed product – with or without a signature from the licensor.   As far as public perception, Dyer’s customers would now no longer be aware of this "Knutsen person" from the label – in fact, the new label “implies” that Dyer owns the patent.  Nor would the public find any other maker’s harp guitar labeled “Symphony” from here on out.  I can certainly see Dyer pushing for such a deal, and Knutsen might have stopped worrying or caring after awhile - especially if they sweetened the licensing fee.

None of this quite explains the “patent” inclusion on the Dyer label after 1912.  In fact, Dyer never removed the patent notice from their labels.  It is still there on the highest serial number we have a photographed label from.  This instrument is circa dated at 1922 in the old system, and in any of my new timelines, still approaches 1920.  How did they get away with it?  Well, as Tom explained, a patent holder was allowed by law to leave a patent notice on a product even after its patent had expired – and Knutsen himself did so, stating “Knutsen’s Patents” on the post-1912 “New Hawaiian Family” labels (which included a photo of an older harp guitar).  Dyer must have followed suit, although they were in a much shakier gray area.  A final frustrating fact is that Dyer included the “Patented in U.S. & Canada” on the labels of the very first to very last harp mandolins, even using them even on the later (brand new) mandolas and mandocellos.  Tom Noe has no answer for this one, it seems to be a completely fraudulent claim.

Bottom line: We have two options here, one of which requires a paradigm shift.  Neither is yet provable.  Either the new 600 series Dyer label appeared in February, 1912, or it appeared in the 1906 to 1908 timeframe.  I favor the earlier timeline.

F) The “lower bass point” harp guitars built by both Knutsen and the Larsons

A key question for both Knutsen and Dyer/Larson fans is of course “who influenced who” – did Knutsen come up with this distinctive design, or did the Larsons?

If we use the 1912 600-series switchover timeline, then it’s simple: Knutsen was building only “lower bass point” or “double point” specimens in Seattle , meaning from 1906 or at minimum 1907.  Additionally, many of these Knutsens were short scale (19-20”) harp guitars.  Thus the Larsons would have been copying Knutsen in 1912, modifying the shape every so slightly and using this design specifically for their own short scale (~21”) Type 3 harp guitar.

If we consider the 1906-1908 timeframe for the label / 600-series switchover, then things become a little cloudier.  It would depend on how accurately we could pinpoint both Knutsen’s first instrument of this type and the exact date of Dyer’s new license / label / 600-series.   And neither is yet possible.

Whichever scenario we accept above, for a variety of reasons, I believe that this was a Knutsen design, as was the “double point.”  As presented in the Noe/Most book, and corroborated by myself throughout the Knutsen Archives (as a “working theory” if not fact), I strongly believe that this model was introduced by Knutsen very early – say, mid-1906 at the latest, whereas for the various 1906-1908 Dyer scenarios for their Type 3, we have a very nebulous period of about two years in which to place it.  Completely unconvincing, I know!  There is also the fact that Knutsen continued to build it as his main model until about 1914, while Dyer’s specimens are extremely few and of only low 600-series serial numbers.

Bottom line: I have no proof, only a strong “gut feeling” that the Larsons briefly “copied” the new 1906 Knutsen harp guitars. 

G) The strange case of the “Style 3”

Update, Sept, 2011: In the previous version of this article (June, 2011), this was an extremely lengthy and troublesome discussion of how in the world Dyer’s “Style 3” appears to have occurred a few years after the Style 4 through 8. As it is something Bob Hartman and I thought long and hard on, I have archived it here. 

However, since the discovery (in August, 2011) of Type 1 specimen #125, I think we may be over-analyzing it.

The answer may be simply this: The specimen #125 (refer to my Gregg’s Blogg article) finally showed us that the label’s Style number did not include a “U” or a “0” which we thought we were seeing on the identical specimen 127.  Instead, both turn out to be a Style 5.  Curiously, the plain appointments match the later Type 2’s Style 4, not its Style 5 with top binding. But so what?  Obviously, Dyer/Larsons made a switch – a new ornamentation designation system – when they brought out the Type 2.

My guess?  Simply that when the first Type 1’s were built for Dyer (beginning with the theoretical serial #101), someone chose a simple numbering system that arbitrarily assigned “5” as the lowest number.  Maybe it seemed like a nice round number, who knows?  Remember that they were just getting started and probably hadn’t yet come up with the specifics on how many “models” (of ornamentation) they might do.  As two of the four known Type 1 instruments match the later Style 7 ornamentation, we know that there were at least two “models/styles” (again, just using the original Knutsen Symphony form).

By the time the Type 2 was designed - not too long after – they had probably come up with the idea of the five different “Styles.”  For whatever reason, “4 was the new 5” (to use a modern expression).  Style 4 was now the plain model, and there was nothing below it.  And they were off and running.

When they later decided to do the new “Seattle Knutsen”-like, short scale instrument, they had a dilemma.  This was not a new “style” of ornamentation on the same instrument; it was a whole new “model.”  But, for whatever reason (ability to just use the same labels?), they simply assigned this new shape (which itself occurred in a couple levels of ornamentation) its own “Style” number. They could have picked anything – “3,””1,” “9,” even a word like “Terz,” whatever.  But they chose “3” which happened to be adjacent to their existing lowest number.   

Update, January, 2012: The latest (sixth) Style 3 found is in the family of the original owner, and family lore says it was purchased "about 1909."  That is only 1 year away from 2 of my Timelines below which propose c.1908 for this instrument (No. 610).

Update, January, 2013: A seventh Style 3 has long been hiding right under our noses, owned by Bob Brozman (who tragically took his own life in April, 2013).  Curiously, it has a somewhat later serial number: 673 (not a hard to read "613" but definitely a "7" according to Bob. Without being able to see it, I suggest that it is in fact "613".)

Bottom line: That’s my current theory.  As far as why they chose “3,” who knows (or cares)?  The dating, the timeline – none of it matters for this discussion.  We just have to accept that we were fooled by an apparent “sequence” – Bob and I were looking at “3, 4, 5, 6, 7, 8” and stuck thinking “consecutively,” wondering where Styles 1 and 2 were!  Problem solved…(?)

H) The first harp mandolin ad shows a different instrument than the one produced
I) The appearance of harp mandolins between the expiration dates of the Livermore and Gaskin patents
J) Patent claims on all mandolin family labels

These three topics are related and I’ll discuss them fully in the separate Dyer Harp Mandolin study.  They don’t relate to our discussion of the harp guitar timeline, and provide no real clues that would help us in those theories.

Regarding timeline clues offered by the harp mandolin serial numbers.  There aren’t any.  Unfortunately, as our study progressed with vague evidence, we got stuck on trying to fit the harp guitars, mandolins, and mandolas/mandocellos into one serial number system.  It cannot be.  I’m convinced that the harp mandolins had their own sequence, probably also starting with 101, but in a much later year (somewhere between 1908 and 1910), and then much later the brand new harp mandola and mandocello used completely different systems (a real Pandora’s Box of a mess, discussed on the Dyer Harp Mandolin page).

Which brings us to:

K) Duplicate serial numbers

If I’m right about the separate serial number systems above, then one would logically expect to find duplicated numbers among the three instrument lines.  Do we?  Yes and no.  We only have one example so far, and as is so often the case, we are lacking proof.  Bob lists a harp guitar #264 and also a harp mandolin #264.  Thankfully, the harp guitar owner supplied a photo (at right), but the mandolin owner did not.  So we’re back to speculation.  The owner may have misread it or miswrote it in their communication to Bob long ago.  So as of yet, we don’t have proof of duplicate serial numbers (which would pretty much prove my case).  So obviously, I would not be surprised if these numbers were the same – in fact, I’m surprised no other duplicates have yet turned up.  Give them time.

Bottom line: Dyer mandolin dating is as problematic and frustrating as the harp guitars. Let's just hope we can discover additional clues!

These four lists coincide with the four Timelines shown at the top, and include only harp guitars.  List A is Bob Hartman’s last provided list.  In the other three timelines, I have spread the numbers out very approximately (give or take several months), to try to keep realistic for the Larsons’ build schedule, while taking the advertising campaign into account.  NOTE: These are my own conclusions and "guesstimations."  It could be that the last serial number known (923) is from 1920, a couple years before, or even many years after.  

Now that you’ve examined the different timelines above, and weighed the Pros and Cons of each yourself, here are the options for dating your Dyer instrument.  “Circa” should precede all dates below.  Though I personally favor Timeline D, I would play it safe for now and include the entire range.

DYER HARP GUITAR
Serial Numbers
All harp guitars are the common Type 2, unless otherwise noted
Known or reported duplicates appears as "(2)" or "(2?)"

(Mandolin family instrument serial lists here)

To Harp Guitars