Jenkins ad, 1902 - this "new 14-string" instrument may have preceded the 18-course New York Harwoods, or it could have been a later "improvement."

Evidence points to the latter, as there is a dated 1898 12-bass Harwood.

So - "Harwood" was a brand, not a maker as I had originally thought. Further corroboration: soon after I learned of this, I received the article on the left from friend Kelly Williams. It's from The Music Trades, dated July 26, 1902. It shows what may be the first Harwood harp guitar, though the name is not mentioned. Clues are the last fret marker (on both necks, in this case) - a bone rectangle engraved "HARWOOD" (as seen above in several instruments), the carved bridge that matches some of the known specimens, the joined headstock, and the position of the necks on the body. Lastly, it's advertised as the new guitar from J. W. Jenkins' Sons Music Co (a continuation of the Jenkins Company referred to above?). Interestingly, they don't call it a harp guitar or similar - just a 14-string guitar. The eight bass strings utilize geared tuners, and the two necks are centered on the body. I've yet to see this model in the flesh.

At this point we still have no idea who (individual or factory) built the New York Harwoods.  Frank Ford, who has examined some of the parlor guitars, believes that they are nearly comparable to Martin quality (but not made by Martin).  There are many mandolins known - bowlbacks mostly, but also flatbacks and even a mandolinetto - and many 6-string guitars, most parlor-size, but a few oversize.  Appointments range from plain to full presentation grade.

Above are all the Harwood harp guitar specimens I am aware of.  All extant specimens appear to be made with stunning Brazilian rosewood back and sides.  Amazingly, my specimen has a perfectly flat top, strung with 18 steel strings, yet is only ladder braced.  Perhaps this is helped by the fact that amongst the evenly-spaced braces (~ every 2-1/2"), one is situated directly above the bridge, and the bridge plate itself appears to be a solid thin piece of ? that completely fills the space between the braces and from side to side.  Left to right, 1st, 2nd row:

• 8 sub-basses, 1 soundhole and symmetrical necks, geared tuners, 2 neck logos, from Jenkins ad
• 12 sub-basses, one soundhole and main neck centered over soundhole, bass neck terminates at body, friction tuners, no neck logo, from American Lutherie Journal #29, Spring 1992.
• 12 sub-basses, same headstock and tuners, but two symmetrical soundholes, the main neck centered over the right hole, 2 neck logos, from a dated historical photo (Aeolian Mandolin Orchestra, Guthrie OK, 1898).
• (Seemingly) identical specimen to the previous, except that it has 9 strings on the neck.  The top 3 strings are doubled; though the tuners all match, the 3 extra strings seem a later add on (full details here).
• My specimen, identical to the previous, but only 1 neck logo. 19" wide (presumably the others are also). On these last 3 specimens, note that the main neck is centered over its soundhole, but the other is not (i.e.: the soundholes are symmetrical on the body, but the necks are not quite centered on the body.  However, the strings almost are).  Note that neck logos do not always occur on the same fret.
• An unusual specimen with the same headstock but only 9 sub-basses, and oval soundholes.
• A parlor size (14-1/2" lower bout) specimen, 6-sub-basses, geared tuners.
• An instrument that clearly has a very different headstock, and appears to have a slightly narrower upper bout and body shape, yet seems to have the bone neck logo.  Of course, other makers could have inserted a similar plaque there as well.

Inlaid bone engraved fret marker