Chapter 18 of the book tells of how The Aegre was overwhelmed and capsized during a storm 150 miles WSW of Tahiti. In the following Notes, I discuss Why sailing boats sometimes capsize and what happens next, Why The Aegre turned completely over and then remained upside down, And how we worked out our approximate latitude and longitude after losing our sextant.
What happens when a sailing boat capsizes?
Some readers will have capsized an open sailing dinghy. Balancing the wind in the sail with your body weight, it doesn’t take much to go wrong, for the boat to tip over, throwing you quickly or slowly into the water. The dinghy coming to an inglorious halt, on its side or completely upside down. A mess of ropes and sails, while anything that wasn’t tied in floats away (or sinks). A skilled crew can quickly turn the boat to face into the wind and then pull it back up the right way, often using the weighted or unweighted centreboard for leverage. Then they clamber back aboard, trim the sails and rapidly accelerate away. The water that’s in the boat shoots out the back. Well, that’s the theory. It doesn’t always work quite like that, and many a dinghy crew are towed home by a safety rescue boat, weary, wet and wiser for the experience.
Larger sailing boats with weighted keels or centreboards can capsize, too, as some readers may have experienced, but the process is a bit different. Most mono-hull yachts don’t rely on the crew weight to balance the wind in the sails but instead rely on a combination of their hull shape and a keel beneath the boat containing a weight (usually lead ) to keep them upright. Mon0-hull yachts most commonly capsize, ie roll onto their sides and possibly go completely upside down, due to a combination of wind in the sails and large waves (that are usually breaking), which can combine together to overcome the keel and hull shape working to keep the yacht upright. This can happen when the wind is very strong, the waves are large with their tops breaking, usually combined with the yacht being sailed fast, and steering control being diminished or lost as the boat heels over. In the worst case, the yacht turns sideways in front of a large breaking wave, known as broaching, which then overwhelms it and can roll the yacht over.
Very occasionally, yachts are turned end over end. This happens when a very large steep wave comes up behind the yacht, lifting the back of the yacht, while the nose of the yacht digs into the water ahead, acting like a brake. If the wave at the back continues to push the stern of the yacht up and forward, it can turn the yacht end over end. This is known as ‘pitch-poling’. Perhaps the best-known example is that described by Miles Smeeton in ‘Once is Enough‘, where the yacht Tzu-Hang is pitch-poled approaching Cape Horn.
In most cases, after a yacht capsizes, the weight in the keel will turn the yacht back up the right way. The mast may have been broken during the capsize, so there may be chaos on deck with the broken mast section and rigging swinging around uncontrolled, potentially damaging the yacht and crew.
With any luck, the yacht’s hatches were closed during the capsize, so little or no water got inside the yacht.
Even if a hatch was open, the amount of water that gets in is usually relatively small compared to the overall volume of the yacht. So, for example, the same amount of water that might just be knee deep in the cabin of a 50-ft yacht might completely fill (and sink) a much smaller, say 25-ft yacht.
The Aegre, at just 21’6″, was the size of a large dinghy but decked over, had no external weighted keel and was a hundred and fifty miles downwind from land, far from a yacht club safety rescue boat. We were running before a worsening storm when The Aegre was rolled over. See The Voyage of The Aegre Chapter 18 for a full account of what happened next.
Following are some further thoughts.
Why did The Aegre remain upside down for a short period? Wasn’t she self-righting? By securing pigs of lead as ballast in the bottom of the hull we expected The Aegre would come back upright if she was knocked over by wind and/or wave action. But we never actually tested if this were so. When it happened, she turned completely over and remained upside down, only turning back up the right way with the weight of Julie and I assisting. We were shocked and surprised.
But in hindsight, I believe I know why this was the case. Much of The Aegre’s stability actually came from her hull shape, relatively little from the lead ballast within her. Fully loaded for ocean voyaging, she was 9″ down on her unloaded waterline, and as such, I believe her cross-sectional shape was that of a moderate displacement vessel. However, her keel was perhaps 2/3 that depicted below. Her stability increased as she heeled over, but as she approached then exceeded 90˚, the righting moment would have started to decrease to the point at, say, 135˚ when the ballast within her would have led to her continuing to turn over until she was completely upside down. At that point, she would have been relatively stable. See the diagram below.
Once upside down, she would have initially resisted turning the right way up again until perhaps 45˚ away from upside down, after which she would have continued to right herself until the right way up again.
This is what seemed to happen. She was initially stable upside down, but then, with the combined weight of Julie and I on one side, she started to roll that way until the critical point was reached, and she continued to roll back up the right way by herself.
Hindsight is a wonderful thing. In fairness, when we bought The Aegre our plan was for coastal cruising, with some expectation we could avoid the worst weather. But then our ambitions grew, one thing leading to another…
Finding the time
Knowing Greenwich Mean Time (Universal Standard Time) was a vital contributor to our eventually finding our way to land, as I describe in the book. Here is a bit more detail.
A few days after the capsize, as we gradually sorted out and cleaned up the cabin of The Aegre, Julie found our chronometer, our accurate Bulova Accutron watch, in the bilges, and still going. I’ve already written about timekeeping aboard The Aegre and how we came by the Accutron. See Measuring Accurate Time Aboard The Aegre.
Our beloved Bulova Accutron watch that we’d bought in Panama City, a predecessor of the quartz watches of today, used a tiny oscillating tuning fork to maintain almost exact time. I say almost because even a Bulova Accutron could lose or gain a few seconds over a 24-hour period. But this loss or gain was consistent. So the watch could be set to the exact time by a radio time signal on one day and then checked against the time signal at the same time on succeeding days to determine its Rate, ie how much it was losing or gaining every 24 hours. Knowing the Rate, and how many days since the watch was set at the exact time, the actual correct (GMT) time could be calculated with confidence even when no radio time signal could be received.
For example, the Accutron watch might be set with a radio signal at exactly 12 noon on January 1st. Checks against the WWV radio time signal on days following might show it gaining 2 seconds per day. Later, on say the 21st of January when the Accutron is showing the time as 12 noon, the actual corrected time(GMT) is 12:00.00 – (20 x 0:00.20) = 11:59:20. So the error on 21 January is 40 seconds fast, based on a Rate of + 2 seconds.
Whenever we checked the watch against the WWV time signal, we wrote down the date, time and error on a sheet of paper kept safely with the watch. Now we’d found the watch, but no sign of the error record. We knew the Rate, but without knowing the error, the watch was useless.
But then, screwing up her eyes and thinking hard, Julie remembered the error in the displayed time on the day we left Tahiti, and with the watch’s timekeeping being so consistent, we could now calculate the actual GMT. This would be a vital aid to working out our position without a sextant in days to come.
Approximate longitude from sunrise and sunset time
Knowing GMT, we could determine our approximate longitude if the sun was shining by measuring our local sunrise or sunset time. I’d already proved this to myself when we were sailing across the Eastern Pacific.
How does this work? Well, by using our Accutron to measure the precise GMT of our local sunrise or sunset, then looking up the GMT of sunrise or sunset at 0° longitude for that same day, at the same approximate latitude, in our Nautical Almanac (which we had painstakingly dried out after the capsize).
Then, by converting the time difference between the two into minutes and dividing it by 4 (because the sun takes 4 minutes to travel one degree near the Equator). The result is your longitude in degrees. So, for example:
On day 13, on 17th September 1974, our estimated latitude was 20° S. The sun finally dipped below the horizon at 04:25 GMT, according to the corrected time on our Bulova Accutron.
At 0° longitude, 20° South, on 15th September 1974, Sunset was at 17:55 (ref 1974 Nautical Almanac).
So, our sunset was 04:25 +(24:00 – 17:55) = 10 hrs 30 minutes later than sunset at 0° longitude
I.e. (10 x 60) + 30 minutes later = 630 minutes later
Converting these minutes to degrees longitude: 630/4 = 157° 30′ West Longitude.
OK, so it’s not very accurate, but it’s much better than nothing. And don’t tell me it’s too complicated for you. I was working this out while exhausted and in the most dreadful conditions aboard The Aegre. But then my life depended on it; yours probably doesn’t.
I tried this method on the long Pacific passage from Panama to the Marquesas, comparing my results with the position I calculated using early morning and evening star sights taken with the sextant. The results largely agreed – but with a consistent error, I called it The Aegre Error and learnt to adjust for it. This error would have come from a combination of the effect of the height of my eye above sea-level and the effect of the atmosphere causing light to appear to bend around the curvature of the earth. These two variables have opposing effects on the sunrise/sunset time, making the sunrise appear too ‘late/early’ and the sunset too ‘early/late’. But back in the Eastern Pacific, I’d found the error reasonably consistent, so now I could allow for it.
Western navigators worked out how to measure approximate latitude from the angle of the sun at local noon hundreds of years ago. But longitude was much more difficult. The easiest ways required an accurate clock so that seafarers could keep track of the time at 0° longitude (Greenwich), but building a clock that would keep consistent time at sea on the moving deck of a ship proved extremely difficult. Dava Sobel in Longitude tells the story of the first clocks to do so. A wonderful read. See a Wikipedia synopis.
Approximate latitude from star declinations
To determine one’s position, one needs latitude as well as longitude.
As the weather eased a week or so after the capsize (and losing our sextant), I write in the book about estimating our latitude using star declinations. The declination of a star is analogous to latitude (on Earth) and is expressed in degrees, minutes, and seconds of arc.
The declination of bright stars is given in the Nautical Almanac.
So, the Pole star is over the North Pole, and has a Declination of 90° (actually 89° 15’50.8″). The star Antares has a declination of 26° S, so if it is directly overhead, the observer must be at 26°S, but for us, it was reaching its meridian altitude to the south, so we had to be in a latitude to the north of 26° S.
Another bright star, Sirius, has a declination of 16° 42. It was hard to tell if Sirius was going to the north or south of us, indicating that we were probably in a latitude of about 17° S . The latitude of the island of Rarotonga is about 21° South, so it was most likely about 240 miles to the south of us.
You might wonder how I could recognise appropriate stars. Was my knowledge of the stars that good? No, it wasn’t. However, while studying how to take early morning and evening star sights in the Eastern Pacific, I learned to use Volume 1 of the Sight Reduction Tables for Air Navigation for Selected Stars, and using these Tables you can easily find named bright stars in the night sky.
To use these Tables, you first use the annual Nautical Almanac to look up the Greenwich Hour Angle (GHA) for the nearest tabulated time (GMT date and time) for that day, then subtract from this one’s estimated angle of longitude; this gives the Local Hour Angle (LHA). Then, going to Volume 1 (above), you find the two pages for your approximate latitude (17° S for us in the table above), then select the LHA in the left-hand column. The columns to the right then show the altitude (angle) and true compass bearing of seven stars suitable for taking a sight on. This makes it easy to find these stars. Luckily the Tables had survived the capsize and drying out.
The basis for our future navigation to land was there in the above. Still, in those early days after the capsize, I wasn’t confident enough in my thinking to actually turn southwest to hopefully find Rarotonga while avoiding the Cook Island atolls lying to the north. I needed to think more and to refine my measurements. So we sailed on westward in what we hoped was the open sea.
Polynesian navigation – all the above, latitude, longitude, GMT and so on, are Western constructs. Long before them, Polynesians were navigating vast distances between tiny islands using natural signs. One time, Doctor, sailor, Polynesian researcher and author David Lewis published research on the direction-finding skills of Polynesian navigators. See Lewis, D 1972, We, the Navigators, The Ancient Art of Landfinding in the Pacific, Reed, Wellington, NZ and also Lewis, D 1978, The Voyaging Stars – Secrets of the Pacific Island Navigators, William Collins Publishers, Sydney. See We the Navigators by David Lewis
As I tell in the book, on the 9th of September, five days after the capsize, we started writing a diary on the only dryish paper we could find, the advertising pages of our Nautical Almanac. If we didn’t survive, we hoped someday someone might find it.
The storm gradually passed, the weather improved, and we built a jury rig. Then with growing confidence in our sextant-less navigation, we sailed on westward, thinking hard about how we could use our knowledge and surviving navigation tables to sail to land.
Next: Notes on Chapter 19: Sailing on to where?
Back to List of Chapter Notes
Back to the Home Page for The Voyage of The Aegre
