I've been checking the accuracy of a 1926 Elgin Grade 478 B. W. Raymond over the past 22 days against computer (NIST?) time. I kept it with me on a holiday trip from Southern California to Colorado and back. I carried it in my pocket through weather ranging from 70 down to 2 degrees and at elevations of 200 to 38,000 feet. During that time, it has lost 63 seconds, easily qualifying it as a railroad grade time keeper.

I would expect it to have consistently lost about 3 seconds a day, but that is not the case. The first 4 or 5 days, it lost about 9 seconds. Over the next couple of days, it came back to about 2 seconds slow. It ran almost dead on for the next week or so. A day or two ago, it was about 80 seconds slow and, right now, it's come back to 63 seconds slow.

Over all, it's keeping railroad time. It would have been certified if it was inspected during the first two weeks but not during the third week.

Is this normal?

If watches wander along the path of accuracy, what keeps the errors from accumulating instead of canceling each other out?

Hey Christopher, If you figure it out let me know. I have a GM Wheeler Elgin 18s,that also wanders about,sometimes fast,sometimes slow,but always within one minute of correct. It can get a minute fast,and I can set it back,and it will gain that minute back,but if I leave it alone,it never gets more than 1 minute off. Just weird. Maybe has to do with temperature,or position,or the magnetic attraction of the galaxy.

I have to wonder if your 478 was affected by the altitude changes it experienced on the trip, as well as temperature shifts.

I carried a Hamilton 992B on a like trip, traveling from South Central Tennessee to Wyoming, over the Continential Divide to Yellowstone, and back. It lost about a minute over a period of 7 days as well. I was very well pleased, and the 992B was a bit further along in technical development than the older Elgin 478. In my personal opinion, I'd say your watch performed perfectly for the type of trip you describe.

In terms of railroad watch inspection, I believe they took the average time of the watch performance over the month to determine the actual accuracy, rather than 'week to week', so yours fell squarely in the approval column.

Regards! Mark

For what it's worth, I'm still timing my 478. After a full month, it is 73 seconds slow. Aside from the 80 second error that occured over a day or two during my trip, the watch has kept to within 10 seconds of NIST time for the entire month. I'm amazed!

If I may offer my two cents based upon my own experiance, dramatic fluctuations in tempurature, altitude, and, most importantly, a change in routine that traveling does.

Your watch did indeed perform great considering. It is amazing how the positions of the watch effect timing consistancy. Sometimes the mere act of laying the watch face up over night zero's out any day time error.

You seem to have a very nice carry watch that can keep up with your lifestyle.

Regards, Matt