An item making the rounds and filling my mailbox: Nature reports on an article in the International Journal of Nautical Archaeology:
A Roman ship found with a lead pipe piercing its hull has mystified archaeologists. Italian researchers now suggest that the pipe was part of an ingenious pumping system, designed to feed on-board fish tanks with a continuous supply of oxygenated water. Their analysis has been published online in the International Journal of Nautical Archaeology.
Historians have assumed that in ancient times fresh fish were eaten close to where they were caught, because without refrigeration they would have rotted during transportation. But if the latest theory is correct, Roman ships could have carried live fish to buyers across the Mediterranean Sea.
The wrecked ship, which dates from the second century AD, was discovered six miles off the coast of Grado in northeastern Italy, in 1986. It was recovered in pieces in 1999 and is now held in the Museum of Underwater Archaeology in Grado. A small trade ship around 16.5 metres long, the vessel was carrying hundreds of vase-like containers that held processed fish, including sardines and salted mackerel.
Carlo Beltrame, a marine archaeologist at the Ca’ Foscari University of Venice in Italy, and his colleagues have been trying to make sense of one bizarre feature of the wreck: a lead pipe near the stern that ends in a hole through the hull. The surviving pipe is 1.3 metres long, and 7–10 centimetres in diameter.
The team concludes that the pipe must have been connected to a piston pump, in which a hand-operated lever moves pistons up and down inside a pair of pipes. One-way valves ensure that water is pushed from one reservoir into another. The Romans had access to such technology, although it hasn’t been seen before on their ships, and the pump itself hasn’t been recovered from the Grado wreck.
Archaeologists have previously suggested that a piston pump could have collected bilge water from the bottom of the boat, emptying it through the hole in the hull. But Beltrame points out that chain pumps — in which buckets attached to a looped chain scooped up bilge water and tipped it over the side — were much safer and commonly used for this purpose in ancient times. “No seaman would have drilled a hole in the keel, creating a potential way for water to enter the hull, unless there was a very powerful reason to do so,” he writes.
Another possible use is to pump sea water into the boat, to wash the decks or fight fires. A similar system was used on Horatio Nelson’s flagship, HMS Victory, in the eighteenth and nineteenth centuries. But Beltrame and his colleagues argue that the Grado wreck wasn’t big enough to make this worthwhile. They say that the ship’s involvement in the fish trade suggests a very different purpose for the pump — to supply a fish tank.
The researchers calculate that a ship the size of the Grado wreck could have held a tank containing around 4 cubic metres of water. This could have housed 200 kilograms of live fish, such as sea bass or sea bream. To keep the fish alive with a constant oxygen supply, the water in the tank would need to be replaced once every half an hour. The researchers estimate that the piston pump could have supported a flow of 252 litres per minute, allowing the water to be replaced in just 16 minutes.
Tracey Rihll, a historian of ancient Greek and Roman technology at Swansea University, UK, cautions that there is no direct evidence for a fish tank. The researchers “dismiss fire-extinguisher and deck-washing functions too easily in my view”, she says. But although no trace of the tank itself remains, Rihll says the pipe could have been used for such a purpose in the ship’s younger days. Literary and archaeological evidence suggests that live fish were indeed transported by the Greeks and Romans “on a small but significant scale”, she adds.
The first-century Roman naturalist Pliny the Elder wrote that parrotfish taken from the Black Sea were transported to the Neopolitan coast, where they were introduced into the sea. And the second- and third-century Greek writer Athenaeus described an enormous ship called the Syracousia, which supposedly had a lead-lined saltwater tank to carry fish for use by the cook.
However, a fish tank on board a small cargo ship such as the Grado wreck might mean that transport of live fish was a routine part of Roman trade, allowing the rich to feast on fish from remote locations or carrying fish shorter distances from farms to local markets.
“It would change completely our idea of the fish market in antiquity,” says Beltrame. “We thought that fish must have been eaten near the harbours where the fishing boats arrived. With this system it could be transported everywhere.”
FWIW, this really doesn’t quite make sense to me, but I’m not in a position to comment because I can’t read the original article, which is horribly, horribly expensive. The Nature piece has a small picture of what the pump + pipes may have looked like and I still don’t get it. The Wikipedia article on maritime hydraulics in antiquity is still ‘in progress’ or whatever and includes this tantalizing concluding paragraph:
This hypothesis has been disputed, since the wooden box protecting a lead pipe along the longitudinal axis of the hull found on the wreck suggest the existence of a bilge well (Oleson and Stein 2007). Previous ships involved in live fish tank transport, entitled navis vivariae, did not employ a hydraulic system, and for this reason the hypothesis proves especially questionable (Boetto 124). For example, one of the wrecks found in Claudius’s harbor at the modern Fiumicino Airport in Italy was a small fish craft dating somewhere around the 2nd century CE. Archaeologists discovered a fish-well in the middle of the ship that used the hole/plump system bored in the bottom of the boat to fill the tank (Boetto 2006). Other than the Grado wreck no other evidence exists of hydraulic systems being used in the fish industry.
… there seems to be a good bibliography there if nothing else, although most items seem to be in the IJNA …
UPDATE (a couple hours later): Thanks to Kristina Killgrove for sending me a copy of the IJNA article, which is full of tech-term goodness. The fish-tank suggestion seems as good as any to me but I can’t help but question the practicality of it. The authors posit a four cubic meter tank on the deck of the ship, with a pump attached to facilitate changing water in the tank. But when would such a tank be practical? A sea-going vessel is going to encounter waves; a tank full of water on deck would be constantly spilling over, no? Even if it had a lid, the pump system would require some way for air to get into the tank for the pump to work, which would diminish the capacity of the tank. Perhaps — if things are as the authors suggest — this was just for specialized situations. Otherwise, I think this one falls into the category of possible, but not probable?
… by the way, Kristina Killgrove uses this article as a point of departure for a rather interesting discussion on the evidence for the Roman diet:Bioarchaeology of Roman Seafood Consumption