William Bentley Fowle (fig. 10), new partner with Hopkins and Colt in the
watch, was born in Boston, Massachusetts on July 27, 1826. His father,
William B. Fowle, Senior, a well-known Boston teacher and educator, had
variously been a bookseller and conductor of a "Female Monitorial
School."[18] The junior William B. Fowle we have first located as a
ticket master with the Boston and Worcester Railroad in 1848,[19] and
he retained this listing in the directory through 1851. Starting in 1852
and continuing through 1862, with no indication of employer or
occupation, he had an office at 9 Merchants Exchange. In 1860 and 1862 he
was a member of the Boston Common Council, and was president of that body
in 1865. In 1862, after the second battle of Bull Run, he raised an
infantry company for the 43rd Massachusetts Volunteers and was mustered
in, September 24, 1862, with the rank of captain. From December 7, 1862,
to March 4, 1863, he was commandant of the military post at Beaufort,
North Carolina. He then reported to his regiment. On June 24, 1863, he
was left sick at New Bern, North Carolina, by his company bound for
Fortress Monroe. On July 21 he rejoined his company at Boston,
Massachusetts, in time to be mustered out on July 30 at the expiration of
his nine months' enlistment.[20]
[Illustration: Figure 11.--THE TWO LEVER ESCAPEMENTS USED IN THE
AUBURNDALE ROTARY. Note, in addition to the escapement, the absence of
banking pins and the metal balance jewel in the escapement at the left,
which is from watch No. 176. (Both watches in the author's collection.)]
In the 1864 _Boston directory_ we find him listed as treasurer of the
Bear Valley Coal Co., and the North Mountain Coal Co., with an office at
38 City Exchange. This association with the coal business continued with
changes unimportant to our story through the directories until 1877, in
which year the name is dropped from the _Boston directory_, not to
reappear until the directory of 1880, where he is listed at "Herald
Building, watches and timers." This was apparently the sales office. The
_Newton directory_ of 1877 drops its previous listing of coal after Mr.
Fowle's name and first mentions the Auburndale Watch Co.[21] In 1866
Mr. Fowle established his home, Tanglewood, in Auburndale, a village in
Newton not far from his boyhood home at West Newton and on the bank of
the Charles River about two miles upstream from the Waltham Watch Co. He
served the town of Newton as selectman from 1869 through 1871, was an
alderman in 1877, and mayor in 1878 and 1879.[22]
[Illustration: Figure 12.--A 24-HOUR DIAL for the rotary watch.
(In the author's collection.)]
William Atherton Wales of New York is credited with introducing Mr. Fowle
to the Hopkins watch. No clue has come to light on what connection there
was between Hopkins and Wales, who had been a partner in the large
watch-importing house of Giles, Wales and Co., in New York and later a
large stockholder in the United States Watch Co. of Marion, New Jersey,
which had only ceased operation in 1874. A patent[23] had been issued
to Fayette S. Giles of New York, the leading figure in the United States
Watch Co., for an improvement in stem-winding watches. This had
presumably been available to his company. In this winding mechanism a
crown pinion driven by a clutch on the stem engages with a large ring
gear, having 110 internal teeth, which in turn drives a gear on the
barrel arbor. The author has seen no watch, except the patent
model,[24] containing this device, but the pillar plate of many of the
United States Watch Co. movements were cut out, apparently to receive
this ring gear.
The expense of cutting so many internal teeth in steel seems reason
enough to explain why this patent did not become the basis for all their
stem-wound models. Steel is far more difficult to cut than brass,
resulting in a much greater consumption of time and cutters, both of
which represent money to the manufacturer. In the patent model these
ring-gear teeth have been cut by a milling cutter which did not pass
through the ring and across the face of the teeth. This produced a gear
somewhat resembling an internal bevel gear, one which could have only the
merest contact with its mating pinion. To make a durable gear for this
application it would be necessary to pass the cutter through the ring in
line with the gear axis. This would require a special or, at least,
radically modified gear-cutting machine with a cutter arbor shorter than
the inside diameter of the gear. Into this short space the spindle
bearings and means of driving the spindle would have to be crowded, along
with the cutter. Hopkins faced a problem similar to this in cutting the
ring gear for his watch, except that the brass gear needed for the rotary
watch could be cut far more easily and quickly. This may be the link
which brought Wales and the defunct United States Watch Co. into the
Auburndale picture. Another plausible link between Fowle and Wales
involves a patent[25] Wales received for a pulley. This, the now
familiar device of interlocking conical sections so commonly used in
variable speed V-belt drives, was assigned to G. E. Lincoln of Boston,
Massachusetts. George E. Lincoln was treasurer of the Mammoth Vein
Consolidated Coal Co. at Boston in 1865, with an office adjoining that of
Fowle. In addition he boarded for many years at Auburndale,[26] and he
apparently owned the buildings about to be converted into a watch
factory. Thus we see that Lincoln may very well have been the one who
brought Fowle and Wales together.
[Illustration: Figure 13.--THE AUBURNDALE TIMER with top plate, balance,
and control mechanism removed to show the train. The conventional barrel
has 66 teeth that drive a pinion on the so-called 10-minute staff. This
staff carries on the dial end the pointer, which revolves in 10 minutes,
as indicated on the dial. Also on this staff is an unspoked wheel of 80
driving the center, or minute, staff through a pinion of 8. In addition
to the sweep hand (or hands in the case of the split model) indicating
seconds up to a duration of one minute, there is a wheel of 80 driving a
pinion of 8 on an intermediate staff. A wheel of 60 on this staff drives
a pinion of 10 on the escape-wheel staff. A pointer on this last staff
also carries the hand that indicates fractions of a second. (In the
author's collection.)]
William B. Fowle's cash book shows, on July 14, 1876, payment to Geo. E.
Lincoln "For large building used $200" and "For small building used $30."
On July 21 is an entry "Milo Lucas bal. of Building Contract $1605.25."
These with an entry on the preceding June 30, "Milo Lucas on a/c Contract
for Building" seem, with a July 25 entry "W. E. C. Fowler, Painting
Factory $64.91," to cover the expense for the bare factory. The
buildings, two stories high and measuring 40 x 20 and 32 x 20 feet,
respectively, were located on the Weston bank of the Charles River,
opposite Fowle's home, from which they could be reached by a private
ferry. This pleasant bucolic location was not far upstream from that
originally sought by the Boston Watch Co. when that firm was looking for
a spot to move to from Roxbury in 1854. The situation of the factory was
described as a wild and secluded glen.[27]
Another account[28] says:
The well appointed little steamer _White Swan_, owned and commanded
by a Captain Gibbs, veteran of the last war, now plies regularly
between Waltham and Auburndale Bridge, carrying picnic parties,
etc.... Along the banks of the river are located the summer
residences of Messrs. Cutter and Merrill, the elegant residence of
R. M. Pulsifer, Mayor of Newton, the splendid mansion of Ex-Mayor
Fowle, the Benyon mansion and others.... At sunset the river is
alive with canoes, row-boats, shells and sailboats filled with
ladies and gentlemen adding, with the delightful music, greatly to
the natural charm of the scenery.
[Illustration: Figure 14.--ESCAPE WHEEL AND PALLETS of an Auburndale
timer. With four pins in the escape wheel, this particular one beats
eighths of a second. (In author's collection.)]
This idyllic pastoral setting surely must have been a joy to all
connected with the little watch factory. It seems to typify the
atmosphere of wealth and leisure into which the infant industry was
brought without adequate study of the problems it would be called upon to
surmount.
The Auburndale machinery came from the United States Watch Co. factory at
Marion, New Jersey, which, as we have seen, was closed in 1874. William
A. Wales, who was associated with Fowle in the Auburndale "adventure,"
had been secretary, treasurer, and director of this company. Most of the
machinery came from George E. Hart and Co., of Newark, which had taken
over much of the Company's equipment, eventually selling it to other
factories. Warren E. Ray, a neighbor of Mr. Fowle's, commenced as manager
of the factory in July 1876, and died suddenly of heart disease about
October 1 of that year. He was soon succeeded by Mr. James H. Gerry, who
had gone from Waltham to Newark in 1863 to superintend the building of
the original machines for the United States factory.
The employees were chiefly drawn from other factories, principally the
neighboring American Watch Co. at Waltham, and the defunct United States
Watch Co., while some who needed no specific watchmaking skills perhaps
never had worked in a watch factory before. Names, not already mentioned,
that have been preserved are: George H. Bourne, L. C. Brown, Abraham
Craig, Frederick H. Eaves, Henry B. Fowle, Benjamin F. Gerry, William H.
Guest, Jose Guinan, Sadie Hewes, Isaac Kilduff (the watchman), Justin
Hinds, E. Moebus, James O'Connell (the stationary engineer), Edwin H.
Perry, Frank N. Robbins, John Rose, Thomas W. Shephard, William H. A.
Simmons, Alfred Simpson, Thomas Steele, Oscar L. Strout, and George Wood.
These, compiled from several sources,[29] represent only a few of the
men who contributed their knowledge and skill to the enterprise; they are
listed in alphabetical order because it has been found impossible to
arrange them accurately according to position, magnitude of contribution,
or length of service.
Of the five Hopkins patents[30] the first and the last are the ones
covering the essential elements used in the Auburndale product. The two
patents assigned in half to William D. Colt apparently were never used,
nor does the device shown in figure 6 seem to have been used, although
these unused patents are listed on the Auburndale movements. Now that the
watch was in the hands of men accustomed to making watches, some
modifications dictated by their experience and by considerations of
expediency in manufacture were made. The movement that issued was 18
size, rather thick, cased at the factory in a nickel case made by the
Thiery Watch Case Co. of Boston, Massachusetts. In the winding and
setting mechanisms, some changes in details were made with respect to
those shown in figure 7. The dial is mounted by means of a rim which
snaps over the edge of the movement as on a high-grade Swiss watch of the
same era. The usual dial feet, if used, would have interfered with the
rotation of the movement. For the same reason, of course, there is no
dial indicating seconds.
[Illustration: Figure 15.--VERGE AND LEVER for an Auburndale timer. The
one on the left beats eighths of a second; that on the right beats
quarters. (In author's collection.)]
Five jewels are found in most instances, two cap jewels and two hole
jewels for the balance staff and a jeweled impulse pin. One of the faults
of the movement is that the cap and hole jewels on the balance are not
separable for cleaning. After the jewels were inserted part of the
setting was spun down over them, making the assembly permanent. A few
movements with only one jewel are known, the cap and hole jewels being
metal "jewels" likewise set under a spun-over rim. Whether or not the
impulse jewel found in these last-mentioned movements is original or a
later intrusion remains undetermined. It is easy to conceive that the
factory would see no more necessity for an impulse jewel than for other
jewels.
The lever escapement is the only one known to have been used, but two
varieties of this are found (see fig. 11). One is a standard club-tooth
lever with banking pins, the other, much more interesting because
unconventional, has pointed pallets and all the lift on the escape wheel,
which has very short stubby teeth, very much like the wheel of a
pin-pallet escapement. No banking pins are used, the banking taking place
between the pallets and the wheel. An examination of a number of these
watches, with serial numbers ranging from 46 to 507,[31] reveals no
correlation between the serial number and the style of escapement, from
which one may conclude that the pointed pallet escapement was originally
used; later four balance jewels were added and the escapement changed to
the conventional club-tooth pattern. As complaints came in about the
defective running of the watch these changes were apparently substituted
at the factory in customers' watches. The movements with the
pointed-pallet escapement seldom show much wear; on the other hand, watch
no. 224,[32] which has the conventional escapement and five jewels, is
very much worn and must have run for many years.
These watches are stem wound by turning opposite to the usual direction
and are set through the winding crown after actuating a setting lever
located under the front bezel. The plates, bridges, and ring gear are
nickel-plated and highly buffed, making a very showy movement, the only
instances of such a finish on watches in the author's experience. In
figure 12 is shown a 24-hour dial to fit the movement. Special dial
gearing would be required for the hour hand to accompany this dial.
[Illustration: Figure 16.--DIAL FOR 1/10-SECOND MODEL Auburndale timer.
(In author's collection.)]
The first of these watches were placed on the market in 1877, priced at
$10.00 to the trade. Soon complaints came in that they were defective in
operation and many were returned. We have seen from the specimens
examined that there seems to have been no established model produced in
quantity. The dial and the number of jewels varied, as well as the
escapement, suggesting that the owners were groping for a salable variant
of the design for which they had tooled the factory. Probably the pointed
pallet escapement was used first, it being the less expensive of the two.
In addition to the saving effected by not requiring banking pins, the
escape wheel was much cheaper to cut, as the teeth were very short and
strong (see fig. 11). Since the banking took place between the pallets
and the escape wheel, there was no adjustment for the amount of slide;
and since the watches were not made to close tolerances, the slide was
necessarily excessive and consequently power consuming. The conventional
club-tooth escapement was probably substituted as less troublesome,
although the banking pins were fixed and could only be adjusted by
bending them. The pallets remained solid steel, without adjustable stone
inserts.
At this stage of affairs approximately $140,000 had been invested in the
venture, the market was already glutted with conventional watches which
enjoyed the confidence of retailers, and the Auburndale Rotary had won a
bad reputation. The success of any watch depends largely on the
confidence the retail dealers have in it. They are looking for a product
easy to sell at an attractive profit as well as one that will stay sold
and create a satisfied customer. Fowle was of course very much
disappointed; before going into the venture he had been advised that he
could expect to produce 200 watches per day on an expenditure of
$16,000.[33] The watches reached the market at a time, the fall of
1877, almost coincidental with application by D. Azro A. Buck for patents
on what was to become the Waterbury rotary. These patents represented a
new and economically sound expression of the basic ideas of Hopkins. The
Waterbury associates immediately commenced work aimed at getting their
watch on the market by June 1878.[34] News of this certainly reached
Auburndale where they were not only well aware of the cost of producing
their rotary but were also aware of the strict cost and performance
studies which Locke and Merritt would apply to any watch before they
would invest in it. Knowledge of this very able and well organized rival,
coupled with the troubles experienced in manufacturing and selling the
Auburndale Rotary, seem to account for the decision to abandon it. It was
unfortunate that the timing of events happened just as it did for a
little more work on the Auburndale and the tools for making it would
probably have placed it on a firm footing in the trade, although
obviously it could never compete with what eventually became the
low-priced watch, really a scaled-down alarm clock minus the alarm
mechanism.
It is said that about one thousand of the "Rotaries" were made. The
highest serial number to come to the author's attention, 507, may
indicate that only a part of the watches started were finished.
Accounts agree[35] that the next product of the factory was a "Timer"
containing a novel escapement patented on May 28, 1878,[36] by William
A. Wales. Early specimens are marked "Pat. Applied For," but one with the
serial number 996[37] bears no reference at all to a patent, presumably
because issuance of the patent or patents was imminent. Apparently the
timer was in full production before the patent was issued on May 28.
Specimens with higher serial numbers are stamped with three patent dates,
May 28, 1878,[38] June 24, 1879, and September 30, 1879, as seen in
figure 13, which also shows the arrangement of the train. In this
escapement the escape wheel (fig. 14) carries in the rim any suitable
number of steel pins all on the same radius from, and parallel to, the
axis of wheel rotation. In all cases the wheel makes one revolution per
second. The verge (figs. 14 and 15) is so proportioned that the distance
between the points of repose on the entrance and exit pallets will stop
the wheel at intervals equal to half the angular distance between the
pins.
In other words, with two pins in the escape wheel the escapement will
beat quarters of a second, because starting from a point of repose the
wheel will be arrested on the other point of repose after turning through
90 deg.. With four pins in the escape wheel and a suitably proportioned verge
the escape wheel advances in steps of 45 deg. and beats eighths of a second.
The growing trend in this period to standardize the timing of sporting
events in intervals of fifths of a second is reflected in still another
model having five pins in the escape wheel and beating tenths of a
second. By the nature of the verge in this escapement, it will be seen
that the number of beats must be twice the number of pins in the escape
wheel, leaving no way to secure an odd number of beats per second, hence
the 1/10-second model. This must have been a less desirable form because
of the much smaller verge required, plus the problem of accelerating so
much mass from a dead stop 600 times per minute.
[Illustration: Figure 17.--A TIMER DIAL that is probably either
experimental or very early. Note that the fractions of a second
(quarters) are shown on the outside dial instead of on a separate dial.
This dial was converted at the factory for use as the base of a
hairspring vibrating stand. A dial different from this but having the
same arrangement of circles is known. (In author's collection.)]
Figure 16 illustrates a dial for this 1/10-second model which the author
found in a lot of unused parts left over when the factory closed. The
watch had an 18-size 3/4-plate movement with grained nickel finish. The
escapement is special, as we have seen, but the fork, roller, and balance
action are conventional. There are five jewels, four to support the
balance staff and an impulse jewel. The barrel arbor comes through the
top plate with a square, as in a keywind watch, but is fitted with a
winding handle, so that a key is unnecessary. This handle appears to be
an afterthought, because on the earlier models (those with serial numbers
below 1,000), the barrel arbor is short, barely long enough to attach the
winding handle; later this arbor was made longer. Patent 204274 issued to
Benjamin Wormelle of Brighton, Massachusetts, on May 28, 1878, the same
date as Wales' escapement patent, may have suggested this winding handle.
On watches with higher serial numbers, there are two arrows on the handle
to show the direction to wind.
The earliest of these timers had a slide on the side of the case to stop
the movement by means of a piece of thin spring steel applied roughly
tangentially to the smooth rim of the three-arm, solid steel balance
wheel. When this action is reversed to start the movement, the spring, in
retracting from the wheel rim, starts the wheel swinging. Soon this slide
on the case was dispensed with by fitting a curved sheet-metal rack into
a groove turned in the edge of the balance cock. Engaging this rack was a
pinion with a square hole through which the square stem could slide to
set the hands back to zero as it had from the beginning, while turning
the stem now would actuate the pinion and rack to start and stop the
movement, as the slide in the case had originally done.
Various minor changes, dictated by experience and the need for economy in
manufacture, were made in these movements. After about the first thousand
the diameter of the balance was reduced from approximately .700 to about
.530 inch. This smaller wheel was, of course, much more suitable to
vibrate at the faster speeds required on the models beating eighths and
tenths of a second. At some time between the manufacture of watches
bearing serial numbers 3135 and 3622, the formerly separate winding pawl
and spring were combined into one piece that could be entirely made in a
punch press. Another economy move was to stamp the name and patents in
place of hand engraving. For a long time hand engraving was used,
although stamping had been used from the beginning on the earlier rotary
watch.
The case was very similar to that used on the rotary. The dial, of white
enamel with snap rim fastened by a screw,[39] carried three graduated
circles, an outer circle graduated in seconds up to sixty surrounding two
smaller subsidiary dials. The top one of these smaller dials recorded
minutes elasped up to ten and the lower one recorded fractions of a
second. The same dial was used on movements indicating quarters and
eighths of seconds, all being graduated in eighths. A dial without
provision for indicating the fractions of a second on a separate small
dial may be seen in figure 17. This last has been made into a stand for
hair spring work and is shown with balance and spring just as it came
from the Auburndale factory with balance spring and wheel for a timer
still in place.
[Illustration: Figure 18.--TAG DISPLAYING DIRECTIONS FOR USE OF THE
AUBURNDALE TIMER.
(In author's collection.)]
The sweep second hand and the minute register hand are attached to
heart-shaped cams friction driven from their respective staffs. They are
reset by a bar pivoted beneath the dial and actuated by the stem through
pressure on the crown. An original instruction tag as sent from the
factory with these timers is seen in figure 18.
Figure 19 shows the mechanism of the split-second model as represented
in U. S. patent 220195 of September 30, 1879, issued to William A. Wales
and assigned to William B. Fowle.[40] A split-second mechanism is used
to time the finish of two horses in the same race or any other similar
event. In usual watches of this nature the watch will run along
indefinitely with the extra or split second hand stopped although this
hand will not record more than a difference of one minute from the main
sweep hand. This was not true of the Auburndale, as pointed out in the
instructions. The reason for this is that motion is conveyed to this hand
through a hair spring which would be damaged if allowed to overwind. To
prevent this a stop is interposed which will halt the entire watch unless
directions are followed. The serrated wheel _F_, of hardened steel,
driving the second sweep hand, is cut on the edge with 120 serrations;
stopping of this hand therefore is only to the nearest half second
regardless of how minutely the escapement is dividing time. This is
rather a serious defect as, if timing a horse race as an example, the
time of the fastest horse is taken on this hand which registers a lesser
degree of accuracy than the time recorded on the second and less
important horse. A general view of one of these watches is seen in
figure 20.
-----
[18] _Stimpson's Boston directory_, 1840.
[19] _Adams' new directory of the City of Boston_, 1847-48,
1849-50, 1851.
[20] Records of Veterans Administration, pension application
666 675, National Archives, Washington, D. C.
[21] The _Newton directory_ at this time was issued
biennially on odd numbered years.
[22] S. F. Smith, _History of Newton, Massachusetts_, Boston
1880, p. 833.
[23] U. S. patent 65208, issued May 28, 1867, all rights
assigned to Giles, Wales and Co., March 4, 1867 and
recorded March 8, 1867, at U. S. Patent Office, liber
G9, p. 100.
[24] In the U. S. National Museum, cat. no. 309021.
[25] U. S. patent 179746, issued July 11, 1876.
[26] _Boston directory_, 1865 through 1872.
[27] M. F. Sweitser, _King's handbook of Newton,
Massachusetts_, Boston, Mass., 1889, p. 203.
[28] Smith, _op. cit._ (footnote 22), p. 20.
[29] The sources used were Crossman, _op. cit._
(footnote 8), December 1887; Henry G. Abbot, _Watch
factories of America_, Chicago, 1888, pp. 93-95;
_Newton directory_ for 1875, 1877, 1879, 1881, 1883,
1884-85, and 1885; _Waltham-Watertown directory_ for
1877-78, 1880, 1882, 1884; and William B. Fowle, "Cash
book" (see footnote 14).
[30] U. S. patents 161513, applied for November 13, 1873,
issued March 30, 1875; 165830, applied for July 14,
1875, issued July 20, 1875; 165831, applied for June 9,
1875, issued July 20, 1875; 179019, applied for May 25,
1876, issued June 20, 1876; and 186838, applied for
January 12, 1876, issued January 30, 1877. A French
patent was issued to Hopkins on September 12, 1876, and
a Belgian patent on September 30, 1876. For lack of
records neither has been positively identified but
presumably they are for the same device covered in
U. S. patent 179019.
[31] No. 46 courtesy of the late C. A. Ilbert (this watch is
now in the Science Museum, South Kensington, London);
124, 176, 224, 241 in the author's collection; 161
Abbot, _op cit._ (footnote 29); 250 Henry Ford Museum,
Dearborn, Michigan; 361 F. Earl Hackett; 387 Dr. Alfred
G. Cossidente; 403 Dr. W. B. Stephens; 423 Crossman, _op
cit._ (footnote 8); and an unnumbered movement
illustrated in _American Jeweler_, December 1898, vol.
17, no. 12, p. 371.
[32] In the author's collection.
[33] Crossman, _op. cit._ (footnote 8), December 1887,
p. 400.
[34] Crossman, _op. cit._ (footnote 8), January 1888,
p. 33.
[35] Crossman, _op. cit._ (footnote 8), January 1888,
pp. 400-401; Abbot, _op. cit._ (footnote 29).
[36] U. S. patent 204400.
[37] U. S. National Museum cat. no. 248691.
[38] U. S. patent 204400. The text of this patent speaks of
dividing the second into "halves, quarters, eighths,
etc." and in the summation of claims of "an escape
wheel, _A_, provided with one or more pairs of pins..."
showing that measuring tenths of a second with a
five-pin escape wheel was not conceived at this time. It
is interesting to note that in referring to the drawings
shown in figure 12 the text states "In the present
instance two pairs of pins are used to denote quarter
seconds." Only one pair of pins is shown, which is
correct. This seems, however, to reflect carelessness on
the part of patent attorneys and examiners, as the error
exists in the original manuscript patent application
preserved in the National Archives, Washington, D. C.
[39] U. S. patent 216917, issued to William A. Wales and
assigned to William B. Fowle, was applied for on
November 1, 1878, after the device was already in use on
earlier specimens of these watches.
[40] The mechanism was also covered by British patent 3893,
issued September 27, 1879, to Philip Syng Justice on
behalf of William B. Fowle.