Fitzjames' Demonic Companion

This enlargement from the engraving published in the Illustrated London News seems to show a grotesque face peering out from James Fitzjames' bedside table. The item is a type of candle lantern sometimes described as a 'chamberstick with storm glass'. I've looked at thousands of pictures of candlesticks and lanterns on the web but so far I have been unable to find a picture of anything similar. I've found a few examples with a pierced foliage pattern on the gallery supporting the glass but nothing resembling a face. Perhaps it was a souvenir of Fitzjames' travels, or maybe even it was never really there and is just a private joke by the engraver.

Pemmican

I have long been familiar with the name of Pemmican as the legendary food of Arctic travellers, but I was always curious as to what it was and how it tasted.

So a while back, in the spirit of learning by replication, I decided to make some using the recipe described in Sir John Richardson's 1851 boat-voyage journal.

Not having a malt kiln, I used a domestic fan oven and ground the dessicated meat in a hand cranked coffee grinder. Beef dripping was the fat added.

The cooking smells and the morsels which I consumed, in the interests of science, during the preparation were delicious but, of the finished product, I find myself completely in agreement with Charles Francis Hall's comment in his 'Life with the Esquimaux' when he says:

"This article is eaten not because it tastes good, for it does not, but to live.
It is almost like eating tallow candles."

It certainly more palatable when made into Hoosh, the staple of Antarctic explorers.

Time we put the tinned food back in the cupboard: Busting the Franklin Expedition myth of 'lead poisoning from tinned food'.

The idea that Franklin's men were poisoned by lead from tinned food has reached almost the status of dogma. For many members of the public it seems to be almost the only 'fact' they know about the expedition.

Owen Beattie and John Geiger's bestselling book 'Frozen in Time' (2004 edition) includes the statement "There is no question of the source of the lead found in the bodies of Franklin's men: it came from the tinned foods". This claim is supported by a paper by Walt Kowal and others published in the Journal of Archaeological Science in March 1991, but, contrary to that claim, the source of the lead was questioned from the very first.

ISOTOPES:

Kowal and his co-authors found that the ratios of the several isotopes of lead in solder from cans found on Beechey Island matched the ratios measured in human remains from Beechey and King William Islands. They concluded from this that the lead in both cases came from the same source. That seems reasonable - indeed it's likely that all the lead on the ships came from the same source.

No measurements were made on contemporary British samples of lead, instead the authors referred to a database of isotope ratios of 98 samples of lead ore from the British Isles published by Stephen Moorbath in 1962.

Having found no credible match, they speculated that:

"...since the isotopic ratio for the Franklin materials is so different from most lead sources in the British Isles, it is quite possible that the lead used in the solder manufacture came from some other nation".

Their argument seems to be, that because all of the other sources of lead which the members of the expedition would have been exposed would have come from British mines, then the only possible explanation for the presence of the unusual exotic lead in the human remains was that it had come from the solder of the tin cans.

Unfortunately mass-spectrometry was still in its infancy in the early sixties, and the isotope measurements published in Moorbath's paper do not bear direct comparison with those made thirty years later. In 1996 Brenda Rohl published isotope ratios of 383 samples of lead ore from the British Isles. Her introduction states:

"The data in pioneering papers such as Moorbath (1962) and Brown (1966) are now too inaccurate to be compared with modern lead isotope data".

MINING:

It is implausable that isotope data could be used to trace the origin of any industrial era lead back to a single mine. Even connecting a sample with a single mining region presents difficulties.

Veins of ore with distinct isotope ratios can occur in close proximity. Equally, veins with indistinuishable isotope signatures can occur many miles apart. A small number of smelters would serve numerous mines within a region, producing metallic lead with an isotope signature averaged from the constituent ores.

Lead is probably the most easily recycled material known to man. At the lead mills ingots of newly smelted lead would be melted together with old lead scrap to manufacture whichever products the market required, such as sheet or pipe or solder.

The isotope signature of a typical sample of manufactured lead would therefore represent an amalgam of both contemporary and previously active mining regions. The average composition of the mixture would change only gradually as established mining regions declined and new ore provinces were exploited.

In 1839, Andrew Ure published statistics showing the relative productivity of the different Lead mining regions at that time. The North Pennines (Durham, Cumberland, and Yorkshire) being the dominant region.

Click on image for source document

The following chart compares the Franklin project data with geological data from Rohl, chosen with reference to the above table, and data from 36 samples from the Northern Pennines, from a 2001 paper by Brett Scaife and others.

Click to enlarge.

Data points from the two Welsh counties are represented by the larger circles to make up for their small number. Derbyshire had been of greater significance in the preceding century hence may deserve greater prominence than Ure's figures suggest due to recycling of old lead. There has been no attempt to distinguish the relative productivity of individual mines within a region, nor to represent measurement confidence intervals.

The main feature to note in the above chart is that all the data points for the Franklin samples are within the locus of the data points for the Northern Pennines.

CONCLUSION:

It is plausible that the isotope signature shared by both the solder and the human remains is simply that of common English lead of the period and that the high levels of lead detected in the human remains could have resulted from exposure to any of a myriad of contemporary sources and certainly not uniquely to the solder of the tin cans.

That members of the expedition suffered from lead exposure is not disputed, but there is no evidence to suggest that "it was the reliance of Franklin's expedition on tinned food that was the root cause".

The supplier of the tinned food, the much vilified Stefan Goldner, is exonerated of any blame in connection with the demise of the Franklin Expedition.

The eminent Food Scientist and Science historian Keith Farrer, OBE (who even has an award named after him) was among the first to question the 'lead poisoning from tin cans' hypothesis.

In a paper published in the Journal of Archaeological Science in 1993, Farrer pointed out, among other things, that only acidic foods such as tomatoes can defeat the cathodic protection conferred on lead by the more reactive tin and iron and that the canned provisions supplied to the expedition included no acidic component. It appears that all published reports of canned food being been found with a high lead content have involved acidic foodstuffs.

Farrer's conclusion regarding the canned food was that "their contribution to the body loads of lead or to any incipient ill health in Franklin's crewmen was trivial".

However no scholarly article in a scientific journal can compete for the public consciousness with a bestseller like 'Frozen in Time'.

So I hope this small effort does something to redress the balance.

01-Aug-2011

Franklin's ships took refrigerators to the Arctic!

It may raise a few eyebrows when I confidently state that, for their fateful voyage to the frozen regions, HM ships Erebus and Terror were almost certainly equipped with refrigerators.

It is also a break with long held consensus, and a volte-face from my previous position, to suggest that the railway locomotive boilers fitted to these ships were modified to boil seawater instead of freshwater.

I'll start by explaining that in the context of early Victorian engineering the word refrigerator was used to refer to what is now called a heat-exchanger. The illustration below, from 1828, shows an example of the 'shell and tube' variety. The hot liquid to be cooled was passed through the array of small tubes while the cooling water flowed around them in the casing. A typical usage was in brewing - to cool the wort down to fermentation temperature after boiling.

Early nineteenth century heat exchanger

WATER:

I began to suspect that the boilers used sea water while thinking about the water consumption of the engines.

Captain Frederick William Beechey made the first proposal to fit auxiliary steam engines to Erebus and Terror with the proviso that they should only be used when there was a flat calm and the ice floes were no longer pressed together by the wind.

Beechey wrote "The openings in the ice are generally of short duration, perhaps for eight or twelve hours only."

Samuel Brees' statistics for the Rennies' engines include the information that the tender's 3.3 tons of water was sufficient for 1.87 hours operation. This equates to 21.2 tons in 12 hours.

The Terror's tanks held 6000 gallons or 26.8 tons so operating the engine for 12 hours would use 80% of the ships water, supposing that the tanks were brim full to begin with.

Alternatively the full capacity of Terror's tanks could sustain the engine for a little over 15 hours.

Fresh water is not difficult to find in an Arctic summer as melting snow forms ponds on ice flows but there would be no guarantee when the weather would allow it to be collected.

If fresh water from the ships' tanks were relied on for the boilers there would be a strong possibility that there would not be enough water on board to fully exploit the rare opportunity of an opening in the ice without endangering the crews lives.

Merely getting so much water out of the tanks in such a short time would be difficult. The ships had numerous separate tanks, most of which held around two tons, with stores piled on top of them. Merely getting at the apertures involved laboriously shifting stores around so that a hose from a force pump could be inserted.

Another useful nugget of information is found in the "Particulars of Steamships", in the National Archives.

The entry for Terror reads:

"Machinery & Boilers complete in every respect
On the 1st trial speed of the Terror was 3.6 knots per hour by Massey's log
with the disadvantage of river water priming the boilers it is expected the speed will be 3 1/2 knots per hour."

National Archives, crown copyright waived for non-commercial use.

The entry for Erebus reads:

"Machinery & boilers Reported to be complete,
on the 1st trial. speed of the "Terror" was 3.6 knots per hour by Massey's log
with the disadvantage of river water priming the boilers, the speed is expected to be 3 1/2 knots per hour"

National Archives, crown copyright waived for non-commercial use.

The Terror's speed trial is mentioned in the entries for both ships, which is perhaps another pointer that the engines were identical in each ship. Erebus, being a foot wider in the beam, would be expected to be slower.

The suggestion of using river water is a clear hint that the ships were equipped to use the water they floated in to feed their boilers. No officer in their right mind would have contemplated contaminating their ship's drinking water tanks with the filth which flowed in "the common sewer of London" - the Thames.


SALT:

I suspect that the widespread belief that the boilers could have only used fresh water is based on the assumption that they would have suffered from excessive corrosion if they had used seawater.

Using seawater certainly did cause greater corrosion. A typical lifespan of a ship's boiler using seawater was only around five years while a similar boiler on land using fresh water could easily last for twenty years. This would not be an issue during the very limited time in which the expedition's engines were expected to operate.

Around the time of the planning of the Franklin Expedition there was lively debate as to whether iron or brass was the best material for the tubes of ships' boilers.

Brass tubes conducted heat better and were resistant to scaling while iron tubes were more resistant to overheating and avoided galvanic corrosion.

A specification for paddle steamers issued by the Admiralty in January 1844, and another for screw steamers issued in September 1845, both included instructions for the tenders to include alternate costings for tubular boilers with iron tubes and with brass tubes.

More immediately dangerous than corrosion was the problem of increasing salt concentration.

A build-up of salt could cause a boiler to 'prime' or boil over, sending a mixture of water and steam to the cylinders. Worse, a boiler filled with excessively saline water could develop an insulating crust of insoluble calcium salts (essentially plaster of Paris) on the boiler tubes, diminishing the flow of heat to the boiler water and causing the tubes to fail through overheating.

In 1824 Henry Maudslay and Joshua Field had patented a method of changing water in boilers to prevent the deposition of salt and other substances.
In addition to the normal feedwater pumps common to every steam engine, the engine was equipped with what became known as 'brine pumps' which expelled a small quantity of hot supersaline water from the boiler in proportion to the quantity of seawater pumped in, thus allowing the salt concentration to remain constant. To avoid wasting heat, the hot water expelled was passed through a heat-exchanger or 'refrigerator' to transfer its heat energy to the incoming sea water.

The royal yacht Victoria and Albert (launched April 1843) had engines by Maudslay's complete with their system of brine pumps and refrigerators, and that system was also stipulated in both of the Admiralty's specifications mentioned.


CONCLUSION:

Franklin's ships were fitted with the best, most up-to-date technology of the day, including brine pumps and refrigerators (heat-exchangers). They used sea-water in their boilers.

It seems that studies of the Franklin Expedition have suffered from almost as many false trails as the searches for the Expedition itself.

Hopefully this will be the year that Parks Canada find some hard evidence to measure these speculations against.

George Back, saviour of Franklin's first expedition, unwitting poisoner of Franklin's last expedition

As a member of the first Arctic expedition commanded by Captain Franklin, the overland "Journey to the shores of the Polar Sea", Midshipman George Back made two courageous life-saving journeys, which prevented that expedition from becoming a complete tragedy. It is therefore ironic that it seems that Back may have later played an unwitting hand in adding lead poisoning to the misfortunes of the final, fatal, Franklin Expedition.

In my previous post I suggested that Sir Edward Belcher may have been responsible for installing the sub-standard snow-melters in the Erebus and Terror. I now find I have to eat my words as it appears the arrangement was finalised after the then Commander Belcher hauled down his pennant on the Terror on March 24 1836.

Belcher had modified Erebus and Terror in preparation for a mission to relieve a whaling fleet beset in sea ice off Greenland. In the event James Clark Ross conducted that task in the Cove, a hired whaler, so the Admiralty decided to make use of the Terror on a mission to trace the Northern boundary of the North American continent.

George Back was high in the Admiralty's esteem because of his recent success in his overland expedition to the mouth of the Great Fish River in 1833-35. He was promoted to Captain within a month of his return and was appointed to command the Terror on May 13th 1836.

Some rummaging in the National Archives last weekend produced a couple of letters which throw new light on the story of the ships' galley stoves and their attached snow-melters. The letters are both from Captain Back to Charles Wood, First Secretary to the Admiralty.

National Archives, Crown copyright waived for non-commercial use.

May 16th 1836

Sir,
I have the honor to inform you that the fire-hearth or galley apparatus of the Terror is fitted for a crew of 260 men, and in order to save fuel, I would suggest that a fire-hearth or grate of Williams, similar to that now in the Blazer be placed in its stead.

A note scribbled on a corner records the decision:

National Archives, Crown copyright waived for non-commercial use.

May 16. Store-keeper Genl to furnish one of Frazer's firehearths fit for a crew of 60 men

A second letter highlights a problem with the installation:

National Archives, Crown copyright waived for non-commercial use.

May 20th 1836

Sir,
The first Lieutenant of the Terror having represented to me that there is no scuttle over the condenser for the purpose of throwing in snow to melt, I have to request that you will move the Lords Commissioners of the Admiralty to give the necessary directions.

So what does this mean?

Back's suggestion to save fuel by replacing the over-sized galley stove was a good one. However, working under time pressure to get the Terror ready for the Arctic, the dockyard workers made mistakes. Lieutenant Smythe noticed the lack of a scuttle but other mistakes were overlooked and seem to have had unfortunate consequences.

The workmen who installed the ice tank appear to have thought of it only as a condenser for the steam of the cooks coppers and they were unaware of its additional intended duty of melting snow for water.

If, as seems likely, the pipe (shown on the Terror's plans) connecting the coppers and the ice tank was made of lead, it would have been made in Chatham's Lead and Paint Mills which since 1818 had produced all the Navy's lead products.

Clem Rutter, Wikimedia, CC BY-SA 3.0

Brass or Copper would also be a possibility although in the absence of seamless tubing, which would not be patented until June 1838, crafting such a curved pipe in the time available would be a fine test of the coppersmith's art. The end result would necessarily include many soldered joints.

It would therefore seem to be a reasonable explanation for the observed evidence of lead poisoning on the Franklin expedition that the lead came from the steam pipe fitted between 16th and 20th of May 1836 on the Terror, and from its duplicate on the Erebus. The lead being transferred to the ships' potable water by cycles of cooking and snow melting in the ships' coppers. There are of course many possible explanations (tinned food not being among them), but this has the advantage of having as it's cause something we can point to on the ships' plans instead of being dependant on suppositions unsupported by evidence.

So it seems likely that George Back played an unwitting role in the addition of lead poisoning to the list of hardships endured by Franklin's men. It is a final twist of irony that, after the ships were abandoned, the retreating party sought safety by heading for the river which was the source of Back's fame and which bore his name.

Did Edward Belcher poison the Franklin Expedition?

Captain Sir Edward Belcher led the final government sponsored Franklin search expedition of 1852-54.

Les well know is the fact that, during the winter of 1835-36, the then Commander Belcher oversaw the fitting out of Erebus and Terror for service in the polar seas. Years later he reminisced about the Perkins heating system he had installed on the Terror even though, unbeknown to him, it had proved a failure on Back's 1836 voyage and was apparently dismissed as expensive quackery by Sir John Barrow.

I would love to have titled this post "I accuse! - Edward Belcher, with the Lead Piping, at Chatham Dockyard" but as I am still sifting the evidence of changes after Belcher's involvement we have to give him the benefit of the doubt.

The above image is my interpretation of the arrangement of galley stove (colloquially "the coppers") and ice tank installed in 1836, mostly based on details from the 1837 profile of HMS Terror in the National Maritime Museum. I have assumed four boilers (also "the coppers") each with a separate drain cock although the original drawing shows only one over-large tap at the back of the stove. I think it likely that this arrangement continued in use on both Ross and Crozier's 1839-43 Antarctic expedition and the fatal 1845 Franklin Expedition.

William Battersby originated the suggestion that the source of lead poisoning on the Franklin Expedition was the water making system installed on the ships. I consider his suggestion of lead for the material of the connecting pipe (shown in red) between the Fraser stove and the overhead ice-tank to be of particular significance.

As far as I know, the only account of this system in use is found in the journal of Royal Marines Sergeant William Cunningham of HMS Terror. It is instructive to compare some entries for 1841 in that journal with corresponding accounts in the ship's log.

5 Feb: Thawing Ice in the Coppers all day and night:

[7 Feb: Rec'd in the tanks 750 galls of water from the ice]

13 Feb: Large pieces of Ice falling from the Ropes aloft which gathered up and put in the Ice Tank for Thawing.

[16 Feb: Received 70 Gallons water made from the ice]

The relative quantities and timescales of these two examples suggests to me that the primary means for water making on these ships was not the ice tank but, as on Parry's first voyage, the coppers.

Cunningham's entry for 25 March mentions both the ice tank and the coppers:

Got the ice tank down from over the Cooks Coppers and a funnel Shipped on the scuttle which will carry the Steam off the Lower deck and thereby make it much comfortabler and wholesome for everyone.

By 20 December 1841 they'd discovered an easier way to complete their water:

Made fast to a Floe of ice & filled four tanks with beautifull fresh water ice.

Parry's ships from his second voyage (1821) onwards each had a "simple ingenious and effectual contrivance" which melted snow using waste heat from the stove's flue. Apparently the same apparatus also condensed the waste steam from cooking. The ships of the searching expeditions were also equipped with comparable systems. The plans of the search ships Enterprise and Investigator at the National Maritime Museum show arrangements very similar to those installed on Erebus and Terror with the exception that the ice tank extends further forwards with the stove's flue passing up through it to transmit waste heat to its contents. No steam pipes are shown but unless they were happy to have the galley dripping with condensation then they must have been fitted. The unique feature of the ice tanks of Erebus and Terror was that their only connection to the coppers was by the steam pipe. I would expect them to have been efficient condensers but poor ice-melters.

It is very wasteful of energy to cook with the water continually boiling instead of simmering or as James Fitzjames puts it "just not boiling". It also serves no purpose making significant amounts of steam when reheating canned food to serving temperature.

Considering the extra fuel needed to make sufficient steam and the heat lost to the air, I'd make a rough guess that making water by melting snow in the coppers would require around half the energy needed to make a similar quantity in these ice tanks. So I'd say it would be a logical assumption that on the Franklin Expedition at least some of the water was derived from snow melted in the coppers instead of in the ice tank in order to minimise expenditure of fuel.

Looking again at diagram of the galley it is not difficult to imagine how the water produced in this way could become contaminated with lead.

When the stove is being used for cooking each boiler holds a kettle of provisions which is heated by the boiling water on the bain-marie principle. Excess steam from the cooking rises up the pipe to the ice tank but much of it condenses on the walls of the pipe and runs back down, dissolving lead from the inside of the pipe as it does so and contaminating the boiler water.

When cooking duties are completed, the kettles are removed from the boilers and, oblivious to the possibility of contamination, blocks of snow are loaded into the boilers for melting into water. No steam is generated in water making. The contamination is caused by the repeating cycle of cooking and snow-melting without cleaning out the boilers in between.

The water first drawn from the boilers next to the steam pipe would be the most contaminated while the water from the boilers on the other side would be pure.

The ships which preceded, and those which followed, Erebus and Terror never needed to melt ice in their coppers because their ice tanks produced abundant pure water, at no energy cost, using waste heat. It would not matter if the water in their cooking boilers did become contaminated because that water would never be used for human consumption, simply topped up as it evaporated.

This then is my explanation for the lead poisoning on the Franklin Expedition and my suspicions as to the culprit.

Innocent until proven guilty, of course.

Making water in the frozen seas

The Christmas berg, from p258 MCormick's Voyages of discovery

When reading Robert McCormick's (of HMS Erebus) account of Christmas in the frozen regions which Russell Potter posted here, my interest was piqued by the mention of the twelve tons of ice, "having just been taken on board, from a hummock, to complete our water".

National Archives, Crown copyright waived for non-commercial use

The logs of Erebus and Terror for Ross and Crozier's Antarctic voyage are held by the National Archives and, for a fee, copies can be downloaded from their website. The page for 23 Dec 1841, shown above, includes the following entries:

• 1.40 shortened sail and made fast to a piece of ice. Employed watering.
• 7.45 Cast off from the ice, made all plain sail

Which suggests that loading the 12 tons of ice took six hours.
The water reserves for that day are stated as 19.90 meaning 19 tons and 90 gallons.

The tedious task of transcribing the log entries for water into a spreadsheet resulted in some interesting results:

This chart shows the water reserves of HMS Erebus, the upper line in blue, and HMS Terror, the lower line in red.
The first thing to notice is the different capacity of the two ships. Terror's tanks could hold 6000 gallons or 26.8 tons, Erebus could carry 1000 gallons more for a total of 31 tons. From the 24 December the upper line slopes upward as the 12 tons of ice are melted and added to the tanks over a period of 7 days. The Terror's line shows two smaller and steeper upswings where ice was loaded, but in this case it was processed differently. Terror's log for 30 Dec 1841 includes the entries:

• Short'd sail, lowered the topsails & secured the ship alongside a piece of ice.
• Erebus fast to the same piece.
• Emp'd filling tanks with ice.

Next to the figure for the water reserve is remarked "The additional quantity of water shown today is from the ice which was put in the tanks". This seems a wise move by Captain Crozier. He didn't encumber his decks with blocks of ice for a week and he didn't waste any fuel melting it.
Both ships dropped anchor in the Falkland Islands with around 12 tons in their tanks.

The horizontal parts of the chart would seem indicate times when, beset in the pack, the ships were melting ice for daily consumption.

Another interesting feature is the daily water usage for HMS Erebus.
While, in the latter part of the cruise, the Terror's water consumption was an unchanging 54 gallons per day the figures for Erebus Erebus show an interesting pattern of one day 85 gallons followed by two or three days at 66 gallons. What ever did they use those extra 19 gallons for ?

All in all this shows a few of the surprising results which can be teased out of old data by applying modern techniques. It is of particular interest if, like me, you share the view, originated by William Battersby, that the lead poisoning experienced on the fatal Franklin Expedition has its origins in the water making systems installed in these ships.