The Origins of the `Isle of the Dead' Benchmark
by John L. Daly
In the attached article, the main focus on the 1841 Ross-Lempriere sea level benchmark was to demonstrate how sea levels had scarcely moved since 1888 when it was first investigated and its exact height measured by Capt. Shortt . But where the mark was struck originally and the circumstances surrounding it is still the subject of some dispute and makes a fascinating story in its own right.
The benchmark is engraved on a rocky natural cliff on a small isle (the `Isle of the Dead') within the harbor of Port Arthur in southeastern Tasmania, an undeveloped harbor which opens directly to the Southern Ocean. The idea for the benchmark came from Capt. Sir James Clark Ross, the renowned British Antarctic explorer and marine scientist, acting in collaboration with Thomas Lempriere, an official of the convict colony at Port Arthur.
Here again is what Ross said about the benchmark in his 1847 book.
"The fixing of solid and well secured marks for the purpose of showing the mean level of the ocean at a given epoch, was suggested by Baron von Humboldt, in a letter to Lord Minto, subsequent to the sailing of the expedition (Ross' own expedition of the `Terror' and `Erebus'), and of which I did not receive any account until our return (to Tasmania) from the Antarctic seas, which is the reason of my not having established a similar mark on the rocks of Kerguelen Island, or some part of the shores of Victoria Land (in Antarctica)."
Having missed that opportunity, he went with Governor Franklin to Port Arthur in 1841 to see Thomas Lempriere who had observed and recorded tidal, astronomical, and meteorological observations over several years. Ross goes on -
"My principal object in visiting Port Arthur was to afford a comparison of our standard barometer with that which had been employed for several years by Mr. Lempriere, the Deputy Assistant Commissary General, in accordance with my instructions, and also to establish a permanent mark at the zero point, or general mean level of the sea as determined by the tidal observations which Mr. Lempriere had conducted with perseverance and exactness for some time: by which means any secular variation in the relative level of the land and sea, which is known to occur on some coasts, might at any future period be detected, and its amount determined.
The point chosen for this purpose was the perpendicular cliff of the small islet off Point Puer (the `Isle of the Dead'), which, being near to the tide register, rendered the operation more simple and exact. The Governor (Sir John Franklin - a naval man), whom I had accompanied on an official visit to the settlement, gave directions to afford Mr. Lempriere every assistance of labourers he required, to have the mark cut deeply in the rock in the exact spot which his tidal observations indicated as the mean level of the ocean."
Explaining why he chose Port Arthur for a mean sea level mark instead of in the Derwent estuary closer to Hobart Town, where his ships `Erebus' and `Terror' were moored, he wrote -
"The tides in the Derwent were too irregular, being influenced greatly by the prevalence of winds outside and the freshes from the interior, so that we could not ascertain with the required degree of exactness the point of mean level."
The emphases have been added to highlight key points. Ross refers to the intended mark as being `mean sea level' or `zero point', no less than 5 times in these short extracts. His intention in respect of the benchmark is therefore clear and unmistakeable. The mark was intended as a mean sea level mark, not a mere tide mark which would only be of use in Port Arthur, nowhere else. The CSIRO did not mention any of Ross' words in a lecture they gave to the Hobart Royal Society in April 2000, and repeatedly used the expression `tide mark' instead of `sea level mark' throughout the lecture.
The exact date on which the Ross-Lempriere benchmark was struck is known - July 1st 1841, because a small stone tablet was placed above it on the clifftop. The tablet went missing around 1913 , but there are two witnesses who reported what was inscribed on it. The first, a Mr Mason  acting for Capt. Shortt, complained the tablet was badly eroded and difficult to read. He quoted its words in 1888 as follows -
"On the rock fronting this stone a line denoting the height of the tide now struck on the 1st July, 1841, mean time, 4h. 44m. p.m.; moon's age, 12 days; height of water in tide gauge 6 ft. 1 in."
Three years later in 1891, a second witness, a yachtsman cruising aboard the yacht `Wanderer' , wrote an account of his voyage in a Melbourne magazine under the pen name `Eight Bells' . He visited the Isle of the Dead and was interested in the numerous gravestones, one of which he sketched and quoted (the gravestone he quoted from is still there today, bearing the exact words he reported) . Then he noticed that -
"...a few yards away, on the rocks fringing the shore and facing the east, was a curious little stone, erected it is said, by Captain Ross, of the `Erebus' and `Terror' Antarctic expedition, during the interval between the visits to the south polar regions. It bears the following inscription:-
`On the rock fronting this stone a line, denoting the height of the tide, was struck on the 1st July, 1841. Mean time, 2.44 p.m. Moon's age, 12 days. Height of water in the tide gauge, 6 ft 1 in.' "
Both quotes from the tablet are presented above exactly as published, word for word, comma for comma. The differences between the two versions are clearly evident. The grammar is different (Mason's `now struck' versus the yachtsman's `was struck'), the time formatting is different (Mason's `4h. 44m. p.m.' versus the yachtsman's `2.44 p.m.') and as is now obvious, the big 2-hour disagreement over the time. Lempriere's tide gauge data for 1841 and 1842, has only recently been discovered in the Royal Society archives in London, and here is a section from his tide log in his own handwriting for July 1841, including July 1st. .
Fig.x Lempriere's tide entries for July 1841. Note the similarity of 5's and 6's
From the data shown here for July 1st 1841, low tide was at 11.28 a.m. with a height of 3 ft 10 in. High tide was at 5.58 pm with a height of 6 ft 4 in. It was sometime between these two tides that the benchmark was struck.
We can infer that mid tide for that day occurred at exactly 2.43 p.m. with a tide height of 5 ft 1 in. These are the exact mid-points between the two times and the two tide heights logged. The reported tide gauge height of 6 ft. 1 in. is only 3 inches short of high tide and could not possibly represent MSL. However, 5ft 1in certainly could be Ross' `zero point of the sea', as this height was not only the mid-tide height for that day, but is also the MSL average for the preceding month of June just prior to the striking of the mark on July 1st. If the yachtsman was right about the time, he merely mis-read the `6' in the `6 ft 1 in' entry on an eroded tablet when it should have been a `5' to give 5 ft 1 in.
Mistaking a `6' for a `5' and vice-versa is a common mistake we all make even today, especially when reading poorly reproduced photocopies and faxes. A 50-year old tablet exposed to the elements would be just as difficult to read as any bad fax copy. In the unlikely event of Mr Mason being right about the time being 4.44 p.m. that would have coincided with a tide of 6 ft 2 in, not 6 ft 1 in. - close, but not exact.
Fig.x High and Low tides for Port Arthur in June/July 1841, with a two-tide smoothing (in red)
This chart of the tides in Port Arthur for the month preceding July 1st shows that 5ft 1in was clearly the closest height to MSL for that month, 6ft 1in being outside the range of possible estimates.
Four factors point to the yachtsman's time of 2.44 p.m. being correct, with a consequent height of 5 ft. 1 in, instead of the 6 ft. 1 in. he actually reported.
1) `2.44 p.m.' reported by the yachtsman is formatted exactly as Lempriere does it in the above log. Mr Mason says it is `4h. 44m. p.m.', not the formatting style used by Lempriere at all.
2) Mason reported difficulty reading the tablet due to erosion. Erosion
could easily make a `5'
look like a a `6' to both observers.
3) Mason's time of 4.44 pm was local mean time, equivalent to 4.55 p.m.
Standard Time. Being only 9 days past the winter solstice deep in the southern hemisphere,
the sun had already set, leaving the Isle in twilight, hardly suitable conditions for survey work.
The yachtsman's time of 2.44 p.m. was mid-afternoon in bright sunlight (confirmed from Lempriere's meteorological log), conditions much better suited to such a precision task.
4) The tide height of 5 ft. 1 in. is not only the height of mean tide
for that day, but is also the mean
sea level one could calculate for the entire month as shown in the graph. It is also the exact height consistent with the time of 2.44 p.m. reported by the yachtsman.
There was clearly no point in striking a benchmark at 6 ft. 1 in. when it does not mark mean sea level (as insisted upon by Ross several times), does not mark high water (which occurred over an hour later), does not mark low water, and indeed does not mark anything at all except an arbitrary height on a tide gauge which no longer exists. A `benchmark' which depends on the continued existence of a local tide gauge for its meaning, is not really benchmarking anything.
Since high or low tide is specific only to the location in which it occurs, only a mean sea level mark, having wider global implications, would have interested Ross, given what he said about Baron Von Humboldt's proposal. Ross and the colonial governor, Sir John Franklin, were both keenly interested in this benchmark precisely because of Humboldt's proposal, and it is inconceivable that the mark would be struck anywhere else other than `zero point' or `mean sea level' as stated very explicitly by Ross several times. Anything less would hardly have rated a mention in his book about his Antarctic voyages.
The conclusion to be drawn from the body of evidence is that the yachtsman got the time right, and that the only possible associated tide height for that time, Ross' `zero point', must therefore be 5ft 1in.
Even allowing for the CSIRO's provisional mid-range GPS estimate of 9.7 cm for land uplift over the 159-year period, there is still have an apparent fall in sea level of just over 25 cm to be accounted for (35 - 9.7 = 25.3 cm). Since most tide gauges record variations in sea level of that magnitude over months and years, as is clearly demonstrated in the tide gauge graphs in the main article, there may not really be any need to account for it at all. It is clear that a general fall in sea level did not happen in the 20th century as other tide gauges around the world (which became more numerous in the 20th century than in the 19th), would have recorded it, so we must look to the 19th century for an answer to the `Isle of the Dead' enigma.
Mr Mason who read the tablet with some difficulty in 1888 was acting for a marine researcher, Capt. Shortt, who was investigating the origins of the benchmark and searching archives in Hobart and Port Arthur for information about it, and without the benefit of Lempriere's tide data. Shortt reported his findings in a short (sic.) paper published by the Royal Society in Hobart .
In order to measure sea level under similar conditions which existed on July 1st 1841, Shortt compared MSL and the benchmark height when the age of the moon was 12 days, as cited on the tablet. His conclusion, in 1888, was that the benchmark was 34 cm above mean sea level, only a centimetre less than the CSIRO estimate of 35 cm 112 years later. Shortt also noted that -
"...it is interesting to place on record, that Captain Miles has learnt from the half-casts in the Furneaux Group they have noticed within the last few years that there seems to be less depth of water over certain well-known rocks near the islands than formerly."
(The Furneaux Islands are half way between Tasmania and the mainland at the eastern end of Bass Strait. Descendants of Tasmanian aborigines were resettled there during the 19th century by the colonial authorities. A general sea level fall at Port Arthur during the 19th century would certainly have been noticed there also as the aborigines frequently engaged in inshore fishing).
Further evidence for a sea level fall comes from period paintings of Port Arthur, all of which show a higher shoreline in the colony precincts than exists today. Indeed, large wooden vessels were unloading cargoes right inside the cove where Port Arthur sits, but today there is insufficient depth of water to do that.
In 1985, Hamon used `hindcasting' of the Hobart data to predict, retrospectively, the state of the tide back in 1841 . Hamon's hindcasting matches Lempriere's data closely, so it can be safely assumed that Lempriere's tide gauge and his data log are also accurate.
Drs John Hunter (CSIRO), David Pugh (Univ. of Southampton, UK) and Richard Coleman (Univ. of Tasmania) jointly authored a confidential internal report for the CSIRO , but have not published anything in the scientific literature. Their recent conclusions, reported in this article, comes from a lecture they jointly delivered in April 2000 to a small audience at the Royal Society in Hobart , a lecture I also attended.
Surprisingly, the Ross written account about the benchmark was not raised once by either of the three speakers. They merely mentioned him as having an idea for a `tide' mark (as distinct from a `sea level' mark) without informing the audience exactly what he had written. When relating the details about the tablet, they informed the audience about Mr Mason's version, but not a word about the yachtsman's version, even though they are familiar with both.
When earlier questioned by the Melbourne `Herald-Sun' on the issue , Hunter claimed that Ross must have `mis-remembered', and in fact carved the mark above mean tide to show where he'd measured the tide in a tide gauge. This scenario would render the whole exercise pointless and contrary to Humboldt's proposal and contrary to the Governor's instructions. When questioned by the BBC,  Pugh said "From all the evidence we know, it was the high water level at that time". He obviously did not regard Ross' own words on the matter as constituting any part of "all the evidence".
Ross was an early pioneer of sea level research, being the first scientist to be credited with calculating a correction factor for sea levels due to variations in atmospheric pressure . Thus he was no mere amateur at the science of sea levels as Hunter and Pugh seem to imply.
In keeping with the lack of publicaton in scientific journals about the benchmark, no written paper was offered at the lecture, the information given here being based on notes taken by myself. Without publication in the scientific literature, the IPCC can choose to ignore even of the existence of the benchmark, let alone integrate it into their assessments of past and present sea level change.
John L. Daly
19th June 2000