Chronometry: Chronometer accuracy

[burt]

I've been writing both in a Bulletin Article and on this board about the Negus brothers and their chronometers. I've also featured Negus #1273 as it has a very impressive history of service in the United States Navy and I also happen to be fortunate enough to own it. It is an instrument that was no doubt accurate enough to be assigned to various Naval ships of war but also to important scientific and geographic expeditions. I was curious as to how accurate these chronometers were back in the day and what results the Naval Observatory achieved in testing them. Here is a U.S.N.O. report from 1912 which records testing results of 1273 when it had been in service nearly 50 years! These are mean daily rates.I think most impressive. I don't know if some of the other results can been read from the document but other chronometers achieved impressive results also and these results are a credit to their respective makers and the Naval Observatory in their efforts to keep these remarkable instruments functioning to the standards they achieved.

January 9-16 +1.46 (seconds)
January 18-25 +0.33
Jan.27-Feb.3 -0.64
February 5-12 -0.06

 

[Jim Haney]

Burt,
To be able to access the records of Negus #1273 is real treat.

I am glad you were able to locate these Observatory records and tract your Chronometer #1273 back to it's testing.

That is a great resource and gives the Chronometer provenance.

I have enjoyed your Bulletin articles on William Bond & Thomas Mercer, keep up the good work.

 

[burt]

Thanks Jim! What I wanted to illustrate here, with information found on my own chronometer, was how accurate these Marine Chronometers were and the great length and precision the Naval Observatory took to adjust,test and document their instruments.I don't think we as collectors have investigated this aspect of Marine Chronometer instruments to any great degree. I'm aware of much that has been written about the Great Observatory Trials and Awards conducted in Europe to judge "specially prepared" precision timepieces in these competitions but this is more of a test of the instruments that were issued for regular marine and scientific purposes. I think for those readers who took the time to look over the single page report much additional information can be found that may answer other questions.The two examples of Sidereal Break Circuit chronometers I found very interesting.

 

[burt]

To put these numbers into some kind of perspective we must start with the fact that the French in 1820 defined that a second is 1/86,400 of the mean solar day. To record a mechanical timing device to the hundredths of that one second required the most skilled mechanical technicians available. When first charged with the responsibility of purchasing,testing and maintaining for issuing marine chronometers the United States Naval Observatory in 1830 had only the most basic instruments. The transit telescope and master clock and the skill of its staff. I've been looking into later instruments that were designed to assist and perhaps speed up the process when the demand for chronometers was escalated by WW II.Not such an easy project as I hoped.

Whitney writes: ( that during the circa 1940's) "After the chronometer has been repaired,it can be brought to time by placing it on a watch timing machine. The Thomas B. Gibbs and Co. made two chronometer timing machines for the United States Naval Observatory. The Hamilton Watch Company also used the same model for rating their chronometers. "
In Mr. William O. Bennett's Bulletin article (put together after his death) writes " The performance specifications for this chronometer (Hamilton 21), established by the U.S.N.O. required a final test of each instrument requiring 30 days. (Back in the old days it was a six month duration trial) A very important part of the final testing system for the chronometer was a sophisticated "time comparator" created by Hamilton for the purpose. I'm not the brightest of persons but here we have two respected individuals talking about two obviously different instruments and neither author mentioning the other?

Information isn't falling out of the trees so I try and look into a explanation for the disparity. I first try and track down Mr. Gibbs. Very little written about him and little to go on but the city of Delavan in Wisconson history records that he started a Company in 1940 which made Timing and Electrical Devices of his own invention. I don't know if he worked at Hamilton prior to that but he filled a U.S. Patent in 1959,while there, along with a Jean Fink for a Clock,motor driven.
As for the "time comparator" that is Mr. Bennett's own invention which he patented while he was working at Hamilton.(U.S. pat. 2,593,650 Applied for July 15,1944) This device permitted the dial error of each chronometer to be determined accurately and speedily each day to the nearest +/- 0.01 second.
In any event remarkable results for the chronometers and of their testing and certainly a tribute to American engineering.Perhaps a bit of a side bar but I think germane to the topic.

 

[Paul Regan]

Burt, while researching my recently acquired Bliss & Co #3119, I was fortunate enough to come across the first Trial it was in at the USNO. In this 1912 Trial it was termed as "new". Further searching produced another Trial in 1917 where it was termed "old". This is a great was to determine the correect age of some of our chronometers. We can also apply other chronometer numbers to approximate the age if we don't get an exact hit. But what really blew my mind was that your Negus #1273 was in the same 1912 Trial as my Bliss #3119. In that trial your Negus is termed "old". So how is that for a cool coincidence!

 

[burt]

Paul,

That is really a "cool coincidence"! What I think also is how impressive the timing results are of your and the other instruments are. You are fortunate to have found a good deal of information to start a pretty good story. You have dated your chronometer and now have proof of its service in the U.S.Navy and that is a good start!
I think we get consumed with so many aspects of these instruments that we fail to realize that whatever they looked at in design or appearance or construction they were built for one purpose and that was to keep excellent rates of time and rate.So why so little interest in this topic?

 

[burt]

Well as the response to my thread is a bit slow and I find the subject interesting I thought I would add to the thread with some chronometer timing rates that I am aware of and give you guys something new to read. As I started the thread with what I feel is a remarkable performance of my Negus #1273 after nearly 50 years of service I thought I'd post some others.

I wrote a story on a Mercer Chronometer #16955,that I own,and that is posted on this board. I have a timing certificate indicating the instrument was keeping time to -0.4 tenths of a second per day when tested on 2-13-1985, while still in service, 41 years after it was built. Impressive I think.

Paul Regan was accommodating enough and provided me with a certificate of his Bliss/Kullberg #3119 certified tested on 6-29-1955 keeping time to -1.0 seconds per day 43 years after being built.

I also wrote an historical article on Bond 199 the oldest chronometer I have records on which was overhauled in 2016 by Dewey Clark. According to the chronometers owner, Richard Schmidt,his instrument was keeping time to +0.72 seconds per day fully 156 years after it left the Bond's shop in 1860 and was purchased by the United States Naval Observatory.

I feel the remarkable results of these chronometers, tested so very many years after being built. gives testament to the builders and persons who maintained them over the years regardless who comments on the thread. I'm including what the C.O.C.S. currently considers "Chronometer accuracy" for mechanical watches that manufactures submit for testing and certification so that their watches can be called, marked and refereed to as "Chronometers". As you can see these limits are well within what every railroad grade pocket watch had to maintain even more than a hundred years ago. What response would a collector receive, on this board if he posted his 992b Hamilton only gained "56 seconds in a week" but is "chronometer accurate"?

 

[Paul Regan]

Thanks for keeping this Thread going Burt. Wouldn't it be nice to have all the known Journals of the major chronometer manufacturers made public via digitization. We all know that the Negus Journals were sold at auction this spring. Where are they now? How many other Journals are out there in "black holes"? I think it would go a long way in instilling new excitement in our hobby(or whatever we call it).
Paul