Possibly First Known Appleton Tracy Pocket Watch

(references [r---] are in my monograph Origins of the Waltham Model 57)

This AT&Co Model 57 #5026, is possibly the earliest surviving Appleton Tracy pocket watch. According to the American Watch Company serial number ledger [r1], the movement was made in July 1857, and according to sales records [r55] & [r102] it was sold July 1857 as 16 jewels with gold balance to Pratt & Reath for $24.25 at 20% discount without a case. (click photos for larger views)


The movement is now encased in an 18K gold hunting case signed G&S with serial number 3026. It appears to be original.

The dial is plain without sunk seconds as were the dials on its predecessor movements, Dennison, Howard & Davis, made by the Boston Watch Company. Actually the dial would be easy is replace because M57 dials are interchangeable, but the number 26 is written on its back. The dial is protected with a hinged crystal cover. The hallmark markings are on both front and back covers; the number is also on the cuvet (movement dust cover). Note gold hands.

G&S is probably maker Gigon & Seydel who made watch cases in Philadelphia, PA, from 1855 to 1859 (according to McElroy's Philadelphia city directory).

Interestingly the case serial number, 3026, almost matches the movement number 5026. One can only wonder if the digits 026 were a coincidence or was the 3 punch mistaken for a 5.


Top Plate and Pillar Plate (under dial) Views

The top plate measures 42.7 mm and the pillar plate measures 45.8 mm, which are consistent Model 57 sizes. I typically get 45.7 mm for the pillar plate, but I often see small variances with these measurements, especially with the pillar plate, which I suppose could be tollarances of manufacture or my ability (or inability) to acurrately make these measurements to 3 significant digits with an old mechanical micrometer. The 3-screw jewel settings are real, but the 3rd wheel jewel is a little smaller, 3.5 mm, than most in predecessor DH&D and later AT movements (4 mm). The wind guard cup is steel as its predecessors; replaced with brass in later movements. The pillar plate is stamped with the subassembly number 26 which is also under the balance bridge; number 6 is under the hour & minute wheels. The comma click, oblong ratchet bridge (over barrel) and long bridge are all standard DH&D/M57 parts. Note the standard "peep" holes in the pillar plate around the pallet arbor jewel through which to see the escapement in motion as in predecessor movements.

Three Views: Underside of Top Plate, Underside of Pillar Plate, Train

The serial number on the underside of the top plate is believed to be the original DH&D plate number. The U-Bridge (Buterfly) potance is the same part on DH&D movements. The center arbor is jeweled giving the movement 16 jewels. The train is standard M57 (and later DH&D) with the 4th wheel pinion above its wheel. Note the 3-spoke pointed tooth (ratchet) escape wheel and English closed pallet as on DH&D movements.

Shown below is a high number DH&D movement compared with the AT 5026.

As can be seen the layouts are identical (and measure the same). The style of the engraving on the barrel bridge is the same. The engraving on the balance bridge is different although similar. The main difference is the placement and font of the signatures. This signature also changed as the M57 evolved.

The DH&D movement is #4887, the highest numbered complete movement I've recorded to date (March 2016) and I have been tracking these watches for over 20 years. Indeed, I believe Howard and Rice absconded with #4891 through #5300 Boston Watch Company movements (or so per shown in table below at end) when the company went insolvent and they lost out at the May 9, 1857 auction. Rice was specifically accused in the insolvency papers [r106] of removing "movements numbered 4891 - 4910 inclusive and other movements."

The BWCo maintained a serialization of matched parts; see [r103]. The serial number is stamped on top of the barrel bridge. All of BWCo movements appear to have their full 4-digit serial number stamped horizontally on the underside of their top plates (because there is no number on top). The last two significant digits (subassembly number) are stamped on the pillar plate and on the underside of the balance bridge. The minute and hours wheels have the last significant digit. This serialization was continued by Royal Robbins when he bought the company at auction.

According to the sales records Robbins' first 20 movements were signed Tracy Baker & Co in name of his investors (5001-5020). After a couple of months Baker had a falling out and Robbins brought in his business partner Appleton, and renamed the company Appleton Tracy & Co. Movements probably starting 5021 were signed Appleton Tracy & Co. As we have seen over many years of watch sales, movement #5012 might be the only surviving TB movement; and #5026 might be the earliest surviving AT movement.

Howard & Rice and Robbins did not (or were not able to) maintain the rigid serialization, though. The visible serial numbers of their movements do not match the numbers inside. I believe the explanation is because both parties were quickly trying to do the best they could with the inventory they had on hand; and I believe many plates were already pre-numbered. The insolvency papers [r106] list a lot of inventory including 100 movements 1/2 to 7/8 done, a long list of parts in various stages of completion, and 1170 pairs of plates! This inventory was incorporated in H&R movements and in early TB/AT movements. It would make sense that H&R would take off with low numbered parts and Robbins was left the high numbered parts. So what we see are matched part numbers inside the movements but different from the visible number on the barrel bridge.

Howard stamped the underside of many (almost all) of his Howard & Rice top plates with five digits, the first digit being 1. I belive the four right most digits are the original BWCo serial number; e.g., as in 14930. It is interesting to note that all of the H&R 4-digit numbers are lower than the AT&Co numbers. I don't know why Howard bothered adding the '1'. The below table of the serialization of observed existing watches supports the above hypothesis.

Serial Number under Top Plate

name of
# on barrel
# under
top plate
name of
# on barrel
# under
top plate
Howard & Rice 6003 14930 TB&Co 5012 5812*
Howard & Rice 6006 14927 AT&Co 5026 5826
Howard & Rice 6011 14983 AT&Co 5042 5842
Howard & Rice 6019 14985 AT&Co 5084 5884
Howard & Rice 6037 15022 AT&Co 5158 5958
Howard & Rice 6047 16** AT&Co 5190 5990
Howard & Rice 6060 15048 AT&Co 5221 5921
Howard & Rice 6065 15043 AT&Co 5332*** 5332
Howard & Rice 6112 14865 AT&Co 5341 5341
Howard & Rice 6128 15217 AT&Co 5404 5404
Howard & Rice 6149 14948 AT&Co 5760 5760
Howard & Rice 6161 14977 AT&Co 5823 5823
Howard & Rice 6214 15115 AT&Co 5826 ?****
Howard & Rice 6217 15116 AT&Co 5901 5901
Howard & Rice 6238 15137 AT&Co 6085 6085
Howard & Rice 6252 15157 AT&Co 6377 6377
Howard & Rice 6264 15163 AT&Co 6534 34*****
Howard & Rice 6300 23****** AT& Co 6590 90
Howard & Rice 6343 17 AT& Co 6669 69
Howard & Rice 6424 94 AT& Co 16447 47
E. Howard & Co 6447 27 AT& Co  21792  92
E. Howard & Co 6482 40       
* I originally recorded this as 5012 by observation at Sotheby's on 10/09/04 under poor light conditions, but now having experience in observing 5826 and 5842 just by looking between the plates with better lighting where the '8' was difficult to distinguish different from '0', I believe this under-top-plate-number should be 5812. This is also consistent with the other movements.
** This is not my observation. Wish I could revisit this watch and confirm the numbers.
*** Looks like around #5300 Robbins ran out of DH&D pre-numbered inventory and started his own matching S/Ns.
**** Just drawing attention to the visible S/N 5826 which is the inside number for 5026, but by now this was Robbins' actual new S/Ns. The pillar plate is numbered '26'.
***** Looks like around #6500 Robbins started putting only the subassembly number under the top plate.
****** Looks like around #6300 Howard started using his BWCo subassembly numbers.