NavList:

Mr Bergman,

Thank you for posting the interesting  logs on 24 January, for the rating of two chronometers in port.   The best estimates that I can come up with of the cumulative errors and the rates that might be expected at noon GMT on 15^th September are  :

Chronometer no. 1969 at noon GMT on 15th September :  Cumulative error estimated as 3 minutes 4 seconds slow compared to GMT.  Rate estimated as about 0.76 seconds going further slow each day.

Chronometer no. 5774 at noon GMT on 15^th September :  Cumulative error estimated as 24 minutes 00 seconds slow compared to GMT.  Rate estimated as about 2.10 seconds going further slow each day.

These estimates come from the three graphs attached here for each chronometer, which I hope fit into the spirit of contemporary methods.    Graph A is the cumulative error for each chronometer at each time ball observation, and was used to estimate the error at noon GMT on 15 September.

The logs do not seem to include any systematic calculations of rate based only on the interval between successive time ball observations.  For interest, I looked at this as graph B attached for each chronometer.

My graph C for both chronometers is a graphical form of the rating calculations as written up in the logs, with a sloping line added to the graph as an attempt to estimate the trend in the rate figure, and therefore to try and estimate the rate at noon GMT on the 15^th.  As you know the logs contain between 9 and 11 rating calculations for each chronometer, over durations of from 4 to 26 days, with these durations being overlapping

My impressions from the calculations in the log are that both chronometers covered by the log had cumulative errors that caused them to read slow compared to GMT.  Both appear to have had rates that were causing the chronometer to run more slow as the days progressed.  I would say the magnitude of this rate on chronometer no. 5774 was 2-3 times the rate of no. 1969, and the rate of each individual chronometer appears to me to have reduced during the time the vessel was in port. 

I see that the GMT of the 1pm ball dropping, as written in the logs, is 5 hours 24 minutes 16.38 seconds (if I am understanding the meaning of the written “38” figures). That seems to equate to a longitude of 66 degrees 4 minutes 5.7 seconds.   Bowditch 1888 (in Frank’s archives) has the longitude of the time ball at Reed’s Point there being 66 degrees 3 minutes 45 seconds.   If it was the case that the time of the ball dropping at Saint John was 1pm by local mean time at the time ball’s own position, then maybe the errors I have suggested above might need changing by a second or so. Alternatively the GMT time of the ball dropping was presumably readily available to mariners and might have been well understood to be 5 hours 24 minutes 16.38 seconds as written in the logs.

I assume that the words written in the log are probably in Finnish? but I thought I could pick out “Aug” and “Sept” written maybe in English.  Did it seem like that to you ?

I found it an interesting exercise, so thank you.

Adrian F

Map of Saint John harbour area 1894:        https://upload.wikimedia.org/wikipedia/commons/1/15/St_John_New_Brunswick_map_1894.jpg