Time Zone Practices and Sea Time

Not open for further replies.

Bill West

I came across some items a few years ago, I wanted to write of them more extensively but that has held up sharing this info too long so I am just going to put out the short, no substantiating citations, version for now.

After reading Dave Gittins’ site I was curious about the amount of incomplete navigation data and tried doing some of my own figuring to learn why things were so uncertain. This lead to buying old navigation books and that brought up a note that Ireland adopted GMT in 1916.... what! they weren’t on GMT all along? Off to the library to search out time zone history. Well to my surprise there was nothing in the plain scientific departments but over in Astrology there was much help. Apparently to determine the horoscope of someone born “on the cusp” they get down to what hour of they day were they born and oh, how do we translate the local time practice to star time? To answer this an amateur, back in the 70’s, started a collection of time zone history first for the US and then for the rest of the world. Correspondents dug out great amounts of history for the local practices in their area. It has all been published on a low key basis and can be found in libraries or bought online. The writer is Doris Doane and the title is “Time Changes“.

The specifics for us are:
France -the government moved to Paris sun time in 1816 but the rest of the country did not follow until 1891. In 1911 GMT was tried but rejected and not finally adopted until 1945. GMT was however practiced in port areas. Paris is 2deg 20min 15sec east longitude so it was 9 min 21 sec ahead of GMT.

Britain -GMT all along from 1880 of course.

Ireland -operated on Dublin sun time from 1880 until Dublin summer time started on May 21, 1916. Then from the end of the first British Summer Time on Oct 1, 1916 it has run on GMT. BST helped war production but introducing it to Ireland in the spring would have been a 1:25 jump ahead, delaying the GMT portion of the change until the fall would be more practical. Dublin is 6deg 20min 16.5sec west longitude so it was 25 min 21.1 sec behind GMT. Sam Halpern alluded to this time practice about a month ago.

Halifax -Atlantic standard & New York -Eastern standard. Adopted in 1883, 4 & 5 hrs behind GMT. Canadian Sanford Fleming proposed the idea of world times zones in 1876, his surveying for the long Canadian Pacific railway probably foretold the sense of it. For those doing navigation calculations, notice that sun time at Ambrose Light Ship is about 5 minute ahead of EST. When you plan an arrival there you need to include the 5 minutes with your day’s mileage/longitude figures as part of getting on to EST.

At sea -I hope readers already know that time zones at sea didn’t become common until the Navy progressed to extensive fleet communication for directing routine activities in 1920-25, as per a 1917 British-French plan. Also two liners introduced 5 day Atlantic crossings in the ‘30s, they would only need 5 one hour changes.

Before WWII everybody used sun time based on their ship’s expected noon position. So Carpathia, Titanic and the Californian had similar times Sunday night/Monday morning but it would be a fluke if they were exactly the same. Two of them did make opposing routine changes that night so 5 adjacent times are contained in our story plus the time choices of the other ships.

The times reported at Cherbourg need to be adjusted according to the reporter’s perspective, local citizen or port staff/passenger. We have two departure times that have been reported, 10 minutes apart. Of course take care that a time could be for the tenders casting off, anchor being hoisted, propellers turning, passing through the harbor entrance/out of the breakwater/ or dropping the pilot. All noticeable times but easily taking 20 minutes in all.

For Queenstown are the reports on GMT, Irish or ship’s time? We need to digress here and consider which of these times would the ship be keeping, navigation convenience could even favor using Daunt Rock’s sun time of -33mins. This branches off to how were the courses run, especially on a great circle. Unfortunately that becomes much more subjective and so it is not included here. And with the Olympic information Sam has posted, I’ve noticed that the 1911 practice doesn’t even seem to be the same as 1912.

"Before WWII everybody used sun time based on their shipí¯Â¿Â½s expected noon position. So Carpathia, Titanic and the Californian had similar times Sunday night/Monday morning but it would be a fluke if they were exactly the same."

I beg to differ. Captains set their clocks as they pleased. Carpathia's clock was set to 47-30W, which had no navigational significance for her. Captain Lord used the old convention of setting his clock to his last noon position. Of the three ships mentioned, Carpathia and Californian certainly kept the same time. Titanic's time may have also have been the same, but let's not revive that old argument.

The starting times from Daunt Rock seem to square with the ships still using GMT. Mark might like to comment.
According to the ISM/White Star "Ships Rules and Uniform Regulations" ---

116. TIME TO BE KEPT.--Seventy-ifth meridian time must be used for time of arrival and departure from Sandy Hook Lightship, Five Fathom Bank Lightship, and other points of arrival and departure in the United States and Canada. Greenwich Mean Time must be used in Abstract Logs after the English or Irish land is made. When passing points and ships at sea, either eastbound or westbound, Greenwich Mean Time, as well as ship's time must be used.

259. SHIP'S TIME.-- The Officer of the Watch will see that the ship's time is changed between the hours of 10 p.m. and 6 a.m., the clocks to be set for Noon before 6 a.m. The Engine Room Clock must at all times agree with the Clock in the Wheelhouse, and must be corrected accordingly.

--David G. Brown
Thank you David Brown for posting that. Departure time at the Daunt Rock LV for westbound voyages departing Queenstown was in GMT. Arrival time at Ambrose Channel LV, formally know as the Sandy Hook Lightship before its name was changed after 1908, was in 75th meridian mean time which today is known as Eastern Standard Time, 5 hour behind GMT. According to what was written in a brochure that Mark Chirnside uncovered for westbound passengers on the Olympic in 1925, passengers were informed that the clocks were to typically go back each night about midnight from 35 to 45 minutes except for the first night out. It said the clock change on the first night out is likely to be considerably more owing to the change from mean time to apparent time.
Regarding time on the Carpathia, Rostron said at the American Inquiry that he received word about the Titanic at 12:35 AM ATS on the 15th. He also said that NY time would have been 10:45 PM on the 14th putting them 1:50 ahead of NY time. This message was acknowledged by the Carpathia at 10:35 PM NY time as recording in the PV of the Mount Temple. According to Harold Cottam, Roston had it in his hands within a couple of minutes after running it to the bridge. This would not be the first time Rostron was mistaken about the difference between ATS and EST or GMT.

Bill West

Dave -so the time from ship to ship is even more individualistic than I was cautioning about. Thanks, that’s good additional help for the newer readers who didn’t know there were any time issues at all.
.......-Californian & Carpathia -that’s some coincidence that Rostron’s habit of 47d30’ W was about the same as Californian’s Sunday noon position of 47d25’ W. The other fluke then is that unlike some ships, neither made a time change that evening. Now if only Smith had the same idiosyncrasies our research would be much easier.

Dave & David -I note that the Daunt GMT instruction is in the context of “Abstract Logs”, do you think that is just referring to the card handed out to the Passengers or are you finding that the specific word is used in some of the bridge record keeping activities too?

Also I notice that the language in the regulation is at least 4 years old. I too worked for a company that had a shelf of regulations. They didn’t always reflect what was being done, keeping the books updated wasn’t nearly as important as keeping the actual work up to date. In periods of change and growth there was no effort spent on them at all. But let’s just take this as an observation, it’s dubious as a counter point.

My main observation about Daunt practices is that the Olympic in 1911 used around 75 to 90 minutes for Thursday night’s change while the Titanic only used 58 minutes. I have a hard time making any mathematical fit out of the rest of that trip without putting the likes of 25 or 33 minutes in to Wednesday night. I appreciate that the other Abstract cards on the web show Daunt times that do fit GMT, a 1925 card specifically says GMT, although that fits Irish practice at that later date anyway. But I’m stuck for what would be the story behind Pitman giving us 58 minutes and so I see the Daunt practice as a question to be checked out. A difference between Bridge time and Passenger time is even a possibility. Sam is already working on time practices in general, I’m just trying to caution that something looks out of the usual here.

Sam -Aren’t Sandy Hook and Ambrose two different ships? This thread leads to some charts, the CP3341C one would be good for 1912. Sandy Hook LV shows in the 1878 chart as 3.5 miles east of Scotland at 40d 26’N 73d 51.5’W. Scotland LV is at 40d 27’N 73d 55’W marking the then main “South channel” and Ambrose LV scales at 40d 28.03’N 73d 50.02’W marking the new channel that you mention as taking over for 1908.
.......-indeed a considerably longer change is to be expected the first night W/B, so why was it only 58 mins on Titanic? I’m not denying having seen the strong evidence of normal practice before and after
April 1912 but the 58 minutes is strong too. Now if you add in 25-33 minutes Wednesday night you get 83-91 which is in the realm of the Olympic trips so we are indeed talking similar Thursday-Friday trip plans. Your Olympic July 1911 calcs specifically show 22.3 knots for the 525 miles made from 2:02pm which back figures to a 92-97 minute planned change (well, there are other problems here however). But for the Titanic trip there seems to be something funny about the way the time change was described.

Explanations we could pursue at a future date for this include the possibility that Pitman missed the 25-33 minutes in preparing his memo and so the Thursday-Friday speed is actually 20.9 rather than 21.4. The speed figure from Cherbourg would have to increase from perhaps 20.5 to 21.2 however. Another approach would be to check Queenstown newspapers for “apparently” local times to compare with onboard times for the various events of each ship’s visit.

While we are mentioning Mr Pitman’s memo, it of course has the known typo of 20.14 for Thursday-Friday’s speed of 21.4. Has it also been noticed that there is 0.6 miles revealed by a carry in the division for that day’s speed? 484 even should be 484.6 miles. In finding 21.4 knots he wants to long divide 484 miles by 22.6 hours. He starts by subtracting 22.6 hrs x 20 kn (=452 miles) but instead of showing a remainder of 32.0 he has 32.6. It has to have come from meaning 484.6 rather than just 484. From that point he continues with subtraction of 22.6 hrs x 1.0 kn (=22.6 miles) and records 10.00 miles as the next remainder. The final remainder is .96 miles. The 0.6 miles is microscopic of course but shows again the frailty of seemingly “source” evidence.


Departure time at the Daunt Rock LV for westbound voyages departing Queenstown was in GMT.


But I’m stuck for what would be the story behind Pitman giving us 58 minutes and so I see the Daunt practice as a question to be checked out.
Departure time at the Daunt Rock LV for Titanic voyage departing Queenstown was 2:20 P.M. Ship's time. This is confirmed by Pitman's Memo giving day's run from Daunt Rk to Noon 12th as 22h 38m--clocks retarded 58m. If departure Daunt Rk had been 2:20 P.M. GMT, day's run would have been 23h 3m.

Bill West-- The IMM/WSL Rulebook quoted was originally in the possession of J. Bruce Ismay and has his signature upon it. The book was updated with "stick-in" notations from time to time. It was eventually entered into the records of the New York Liability hearings where my copy was obtained. Judging from the updating it contains, I trust that the book was current as of April 14, 1912.

-- David G. Brown
Bill: Some light station history from US Coast Guard http://www.uscg.mil/hq/g-cp/history/Lightship_Index.html.

Location & historical notes: New Jersey, established in 1868 to mark the wreck of the SS Scotland. The wreck was removed in 1870 and the station was discontinued. Shipping interests considered the station necessary and therefore it was reestablished in 1874. Moored 3.2 miles and 103 degrees from the Sandy Hook Lighthouse and about 4 1/2 miles westerly from the Ambrose lightship. Used as a reference mark primarily by north-south coastwise traffic using the Sand Hook and Ambrose channels in the approaches to New York Bay. The Scotland radio-beacon was said to have been widely used by commercial aircraft making an approach to Idlewild/JFK airport. The station was replaced by the Scotland Lighted Horn Buoy "S," which was 0.4 miles and 143 degrees from the final lightship station, and was equipped with RACON and passing light. Station was named "Wreck of Scotland" from 1868 through 1891 and was then renamed "Scotland."

Location & historical notes: Ambrose Channel, New York, off the entrance to New York Bay, east and south of Sandy Hook, 6.2 miles and 097 degrees from the Sandy Hook Lighthouse. While acting as a relief lightship for WLV-613, Relief Lightship LV-78 / WAL-505 was rammed and sunk on Ambrose Channel Station on 24 June 1960. The station was replaced by the Ambrose Light Tower. The station was originally named "Sandy Hook" from 1823-1908. Lightships assigned:

1823-1829: "V V "
1829-1838: station discontinued
1839-1854: "W W "
1854-1891: LV-16
1891-1894: LV-48
1894-1908: LV-51
1908-1932: LV-87 / WAL-512
1932-1952: LV-111 / WAL-533
1952-1967: WLV-613

See http://www.uscg.mil/hq/g-cp/history/gifs/LS_27.jpg for photo of Olympic passing LV-87 prior to Titanic disaster.
Capt. Collins:

If that departure time was ATS for the longitude of Daunt Rock then an interesting problem developes. LAN at Daunt Rock on Apr 11 was at 12:34 GMT. Adding a 24 hours 58 minutes to 12:34 GMT gives 13:32 GMT for LAN on Apr 12. LAN occurred for that date and time at longitude 22° 50'W. If we put the ship on a rhumb line from Fastnet (51° 23'N, 9° 36'W) to the 19:00 GMT La Touraine message position (49° 45'N, 23° 38'W) at a longitude of 22° 50'W, we find that the ship would have been about 31 miles E by N of the La Touraine message position at LAN. It obviously does not take the Titanic 5 1/2 hours to cover a distance of 30 miles.

The 58 minute setback was bad data. There is a saying in the data processing field, "Garbage in; garbage out."

Bill: The Olympic's 2nd voyage data was published in George Behe's fine book "Titanic Safety, Speed and Sacrific" p. 30. I already posted some of the information in another thread. The log card data
had "Daunt Rock Light Vessel Abeam, July 13, at 2,02 p.m. Departure." Distances recorded were 525 miles to noon on the 14th, 560 miles to noon on the 15th, 534 miles to noon on the 16th, 526 miles to noon on the 17th, 518 miles to noon on the 18th, 228 miles to Ambrose Channel Light Vessel. "Arrived at 10,8 p.m. Passage 5 days, 13 hours, 6 minutes." "Distance 2891" "Average Speed 21.72 knots."

You can do the math. 5 days, 13 hrs, 6 min back from 10:08 PM July 18th in NY time is precisely 2:02 PM July 13th GMT. Departure time is GMT! The first night out the clocks were adjusted back to go from GMT to ATS so that at LAN the next day the ship's clocks read 12:00. (We already talked about a possible need for a slight correction in the forenoon each day so I don't need to say more about that.) Then each night another adjustment was made for the changing LAN longitude. The final correction came by setting the clock from ATS to EST on the final night of the voyage, and the passage of Ambrose was recorded in EST. All this in accordance with the rules that David Brown posted.

Bill West

Dave -just an additional thought on Rostron. I wonder how he shared out the work between the 2 watches while implementing a 1:50 change out of New York and a 3:10 change at Gibraltar? He not only lengthens or shortens somebody’s work he could really clip their sleep too. It certainly can be done I just wonder how.

David -That’ s good support. My thoughts however would be that it might just be “current” to the extent that something was important or unique enough to issue a sticker for. We’re left to ask why the new name and 2 mile move of the principal New York destination wasn’t important enough to be covered. Nor was making the general writing effective by keeping it in tune with what they were actually using as a destination (because if you let the examples get lax then the people following the book get lax). So it might just be showing that there can be a bit of fuzziness in the best of sources. But do remember that I’m not trying to prove it wasn’t GMT at Daunt, I’m just wanting to add a question mark to note that we don’t have a 100% fit with Pitman’s memo. There’s an old logic concept that finding one line of reasoning to be true without proving why the opposite line of reasoning is false carries a large risk that the finding might have a logic flaw, particularly the infamous “begging the question” flaw. In all, I’m just wanting this issue to be taken as some more fuzziness to think about sometime.

Sam, your lead post -thanks for the USCG link, several previous searches up to Saturday had only found their list of vessels which was weak on station location info. But 4 maps say this was a different location, the link says it was a different ship and it was a completely different route into the harbor. So maybe all that has gone on is that Sandy’s immediate retirement has allowed the web writer to hang his words on the fact that it served the same broad purpose. If Sandy had been carried on for even 6 months overlap he would be forced to used the phrase replaced or retired rather than renamed. As ocean distance is our purpose, we have to consider this to be a different location. So its okay for him to call it a rename but we’ll have to keep in mind that he actually means a different location.

Sam, your next post
-your reply to Captn Collins. It is indeed my point that accepting GMT and the 58 mins together won’t fit the rest of the trip. My problem is which is the bad data? That’s were I’m wanting a to place a question mark in the discussions until we can prove one and disprove the other. I say again, I consider the GMT case to be strong but note that it hasn’t answered Pitman’s case.

Its funny, I used “25-33 minutes” change in my post to be non precise and avoid challenge on the detail of which one while we looked at the generalities. So indeed a 33 min change doesn’t fit the miles made, not that I was actually thinking it has to. But on your approach a 25 min Irish AST at Daunt does seem to fit quite close to the 484 odd miles made to noon Friday. The distance to Smith’s LaTouraine message position would then be fine with either 58+25 or 83+0 minutes.

I only put 99% faith in the telegrams because the less critical the message is the more likely the position could be stale or simplistic (they didn’t even get their CQD position right on the first try), we have problems with time practices, how close are the time stamps and we have typos.

Yes, reasoning out a revised Pitman memo might work, but we need do the whole trip not bits and pieces because the bits keep interfering with each other. We need to pencil those positions on the chart I sent, then adjust the spreadsheet and see if we can find a time change, speed and miles that runs through all of them, at the times indicated and arrives at a position that can be linked all the way on through Sunday, especially with the 44, 44 and 47 time changes making sense. I venture we will end up with several pieces of data to be labeled “bad” before we get it to fit.

-your reply to me about Olympic. Yes, that’s the info I took as given, my purpose was merely to place it for contrasting with the Titanic.

So our several posts haven’t changed that “1911 practice doesn’t even seem to be the same as 1912” is a question that is not fully reconciled with all the available information. But I think it will only be when we find more info that we will be able to move this from a footnote question mark to a solid answer, the old ground isn’t settling it so far.

We weren’t quickly in agreement here which made me realize that we are getting caught on the old problem of arguing details when we each have started from different background premises. We need to evolve to agreeing on broad points before we can settle details and yet it is the details that will prove those broad points. So for now I have to suggest we let ”maybes” sit on the table until we have enough to answer a large portion of them at once. Even then there are times when a 99% conclusion is reversed by being unable to answer one last detail.


If that departure time was ATS for the longitude of Daunt Rock

What I wrote was, "Departure time at the Daunt Rock LV for Titanic voyage departing Queenstown was 2:20 P.M. Ship's time."

At sea, ships's time is set for the longitude at noon. However, when the ship is in port it usually adopts the time of that port. This is for the convenience of working with the port- cargo and passenger schedules etc. Therefore, at Queenstown Titanic time would be that of the port, based on longitude 6° 20' 16.5" west, which is 25 min 21.1 sec behind GMT. Noon 12th long 20° 39'W - 6° 20'W =14.31/15= 58 minutes clock set back.


based on longitude 6° 20' 16.5" west, which is 25 min 21.1 sec behind GMT
Please note correction:

Should read: "based on time meridian 6° 15' west, which is 25 min behind GMT...Noon 12th long 20° 39'W - 6° 15'W =14.4/15= 57 minutes 36 seconds(58 minutes) clock set back.

"The IMM/WSL Rulebook quoted was originally in the possession of J. Bruce Ismay and has his signature upon it. The book was updated with "stick-in" notations from time to time. It was eventually entered into the records of the New York Liability hearings where my copy was obtained. Judging from the updating it contains, I trust that the book was current as of April 14, 1912."

-- David G. Brown


It MIGHT have been current as of 14 April 1912, but the book was most likely not part of Ismay's personal collection. As I recall, he brought it to his deposition to be submitted into court records, which is why it bears his signature. I presume that it came from a WSL office. But bear in mind that Ismay's deposition took place long after the disaster. I would have to check again to see if it was "frozen" from 1912 or not. Yet, if the attorneys accepted it, then they must have felt it was germane to a trial based on 1912 events. The Limitation of Liability trial was based on what regulations WSL had in place, not what changes they had made after the disaster.

The book was small, with a bright red cover. IMM/WSL was embossed in gold on the cover. It was well worn from frequent use and opening. Personally, I see little reason why Ismay would have retained it, with additions, for his own use. It doesn't seem to fit with his personality.

Capt. Collins: The 58 minute setback assumes that Pitman's 2:20 was Dublin Mean Time not GMT. Yet, departure times were always recorded in GMT and arrival times in EST for these westbound voyages. Yet, for argument sake, let's take it that Pitman (or whoever work out his table) used 2:20 DMT instead of GMT.

The time difference between departure and noon on the 12th would be as Pitman wrote 22 hrs 38 min or 22.63 hrs. Now 484/22.63 = 21.38 knots. However, if the 2:20 PM time was GMT then the time difference till noon on the 12th would be 23 hrs 3 min, or 23.05 hours, and 484/23.05 = 21.00 knots. So how to resolve the departure time issue because it obviously affects the average speed of the ship? What I'm looking for is supporting evidence which may help resolve this.

One thing that would help is finding the location for noon on the 12th. You gave a longitude of 20° 39'W. We know from that message to the La Touraine that the ship was reported at 49° 45'N 23° 38'W at 19:00 GMT on the 12th. So let's put the ship on a rhumb from Fastnet to that La Touraine message position at the longitude you used. What I get is 50° 06.1'N 20° 39'W at LAN, which would be at 13:23 GMT, and 117.5 miles from the 19:00 GMT message position. Time difference is 5 hrs 37 min, or 5.62 hrs from LAN to the message position. The speed over ground averages to 117.5/5.62 = 20.92 knots.

Now we know from Ismay's testimony that the ship ran at 70 revolutions leaving Queenstown. The next day it was increased to about 72 rpm. The day after to 75 rpm. Based on tests ran on Olympic after the Titanic disaster, Wilding reported that the ship did 18 knots at 60 rpm and 21.5 knots at 74 rpm. It was also verified at the British Inquiry that Titanic would be doing between 21.75 and 22 knots at 75 rpm. These all line up very nicely when plotted out. At 70 rpm the ship would be making about 20.6 knots. At 72 rpm she would be making about 21.1 knots. To average about 2.4 knots for the 1st day out the ship would have to have been carrying about 73 revolutions from departure to noon on the 12th, which she didn't. But to average about 20.9 to 21.0 knots, the ship need only carry 71 to 72 rpm which is consistent with what was reported and checks with the speed we calculated between noon on the 12th and the 19:00 GMT location.

If you have some other way to sort this out, please feel free to offer it.
I noticed a typographical error in the last long paragraph that I wrote above. The correct sentence should have read, "To average about 21.4 knots for the 1st day out the ship would have to have been carrying about 73 revolutions from departure to noon on the 12th, which she didn't." This ties back to what I said in the second paragraph above where the speed over ground assuming a 2:20 Dublin Mean Time departure calculated out to 484/22.63 = 21.38 knots. I hope this was understood.

One of the possibilities for the 58 minutes in the Pitman memo is that the ship's clock's were set back 25 minutes during the overnight run from Cherbourg to Queenstown so that they would be showing Dublin Mean Time for the arrival into Queenstown. It is possible that what was remembered was that on the night of the 11th the clocks went back 58 minutes for noon on the 12th. The departure time at Daunt Rock by IMM rules would have been recorded in GMT (as we see they were for the voyages of the Olympic) on all log cards for westbound voyages leaving Queenstown. It could very well be that 2:20 PM is what was remembered as abeam Daunt Rock, and the setback time remembered was the change from Dublin time (used throughout Ireland before 1916) to ATS that took place that night. Just a possibility to explain where these numbers may have come from.

In contrast, the other setback times in the Pitman memo really don't make sense. They are both 44 minutes for two consecutive nights. That corresponds precisely to 11.00 degrees of west longitude advancement each day, a travel distance that is not consistent with the reported distances that appeared in the memo. It seems to me that those times were put in as a convenience so that some daily speeds could be calculated. If we were to add the total setback times in the Pitman memo 58 + 44 + 44 = 146 minutes, then add 25 minutes to get the offset from GMT, we find that Local Apparent Noon (LAN) on Apr 14th would have occurred at 14:51 GMT, 2 hours 51 minutes past noon at Greenwich. This corresponds to 42° 45' W longitude for LAN on April 14. If this were true, it would put the ship about 218 miles before the corner which simply does not fit with the total miles of 1549 reported in the memo. It comes out about 92 miles short. To get even close to the wreck site location, which know with certainty these days, the Titanic would have to have traveled 352 miles in the next 11 hours and 40 minutes, or a speed of about 30 knots which is ridiculous. Point of all this is that the setback times in the Pitman memo don't add up to the longitude that the ship was at for LAN on the 14th.

The distances in the memo for the first three days, however, can be verified by other accounts given. I believe we are on solid ground there. The setback times used in the speed calculations in the memo are completely unreliable.
Not open for further replies.