Ships and Navies

[tadamson]

To be honest, Morrison is not the greatest of authors available.  I'd start with Rodgers, the good Admiral may have written a long time ago but he does cover the sources very well.
Try:
Lionel Caisson's "Ships and Seamanship in the Ancient World"
Publisher: The Johns Hopkins University Press; New Ed edition (1 Dec 1995)
ISBN-10: 0801851300
ISBN-13: 978-0801851308

And for Medieval
John Prior's "Geography, Technology, and War: Studies in the Maritime History of the Mediterranean, 649-1571"
Publisher: Cambridge University Press; 1st Paperback Ed edition (14 May 1992) 
ISBN-10: 0521428920
ISBN-13: 978-0521428927

As for the engineering arguments, the killer is that as you add extra banks of oars they have to be longer each time.  On a 'one man per oar' banks beyond 4/5 are too long for one man to actually pull it through the water from a sitting position.

Given the archaeology of Roman boats with multiple men per oar, the textual evidence of 5's fitting in sheds built for 3's  etc I have to admit that I can't see anything above a 4 having a one man per oar design.

[Patrick Waterson]

As for the engineering arguments, the killer is that as you add extra banks of oars they have to be longer each time.  On a 'one man per oar' banks beyond 4/5 are too long for one man to actually pull it through the water from a sitting position.

Given the archaeology of Roman boats with multiple men per oar, the textual evidence of 5's fitting in sheds built for 3's  etc I have to admit that I can't see anything above a 4 having a one man per oar design.

Thanks, Tom.  Rodgers, Caisson and Prior are on my list.

I am not so sure that extra banks of oars have to be longer, provided the extra banks of rowers slope outward as well as upward.

Interestingly enough, it would seem (from Age of the Galley) that the majority of mediaeval galleys were one man per oar versions (albeit multiple rowers sat on the same bench), with multi-man oars being a comparatively late development and suffering from several serious problems that were never entirely resolved, notably rower inefficiency and the fact that losing one man on a bench of three disabled the entire oar (some galley captains refused to sail without a spare man for each multi-man oar).  Given Morrison's estimate that one oarsman per oar is 90% efficient and two oarsmen per oar works out at 67% efficient, one can see why few mediaeval designers ever went beyond three men per oar.

Having had a brief and partial read around the subject, it looks as if the two major geometric constraints on rower effectiveness are the Blade-In-Water line (which must not be too low or too high relative to the shoulders) and the thole line (where the oar pivots - about waist level).   Nothing in between these points seems to matter much, so one could have oarsmen who are almost (but not quite) prone.  This might do much to explain why the Ptolemaic 'forty' seems to have been about 40-45 feet high rather than 120 feet high, and might also make the standard polyreme more comprehensible, besides allowing a quinquereme to fit in a standard trireme shed.

It also would allow one to get away with many more oar banks before oar length becomes a problem (as it evidently had with the 'forty').

Patrick

[Jim Webster]

You might be confusing efficiency and speed.
Yes the efficiency per man falls the more men you have on an oar, but the boat goes faster because there is still more energy being put into the system (even after efficiency losses).

Similarly just going 'out' rather than 'up' doesn't help much, because you are still higher, so you still end up needing to row standing up.
Also you become more unstable because your keel width/hull width in the water doesn't increase, you're just supporting a bigger mushroom on the same sized stalk.

Jim

[Patrick Waterson]

Efficiency is going to enter the equation in the power:weight ratio, which is also (hydrodynamic considerations being equal) the basic determinant of speed.  And while an oar with two or three men on it will indeed transmit more power per oar than a one-man oar, this is of little help if one has significantly fewer oars.  300 men heaving away at 100 three-man oars are indeed putting in more energy per oar than 300 men with 300 one-man oars, but much more of that energy is being wasted and the propulsive power is more like 146 'man units' against 270.  Hence the multi-man-oar ship has to pack as many rowers on half the displacement just to stay competitive.

Going 'out' means one does not have to go 'up' quite so far with each tier of oarsmen, so one can pack in more tiers before one's oarsmen have to adopt compromising positions.  There is another trick one can use, as did many mediaeval galleys, and that is to arrange multi-man benches in herringbone fashion, giving each man his own individual oar - known as a zenzile - as opposed to the unsatisfactory multiple men to an oar arrangement (known as a scaloccio).

Confucius he say size of mushroom matters little provided weight of stalk is sufficient (make of that what you will  ).  One does note an increased beam on the one polyreme for which we have measurements, so it seems that Hellenistic engineers were well aware of how far they could go.

Patrick

[Erpingham]

There is another trick one can use, as did many mediaeval galleys, and that is to arrange multi-man benches in herringbone fashion, giving each man his own individual oar - known as a zenzile - as opposed to the unsatisfactory multiple men to an oar arrangement (known as a scaloccio).


Patrick

It should be pointed out alla scaloccio was so unsatisfactory that it progressively replaced alla zenzile in the second half of the 16th century   However, the primary reason was alla zenzile needed more skilled rowers (which may or may not be significant in classical times).

On multi men per oar, the oar length and consequent weight need to be considered.  The higher the vessel, the longer the oars.  At a certain point there has to be a trade off between oar length and the ability of one man to move that oar at a useful strike rate.

On semi-recumbent oarsmen, I think that the upper body needs to be fairly upright to work an effective stroke.  So would a low angle oarsmen have an efficient, maintainable stroke?  Could he apply his full power or would he have a shortened reach over a sitting oarsman?  I suspect this needs some CAD software to sort out.

[Jim Webster]

Also remember, as tadamson said,  they reused ship sheds so the ships didn't get that much taller or wider
Jim

[Patrick Waterson]

Tom pointed out that quinqueremes were housed in trireme sheds, indicating that dimensions of those types were compatible if not identical.  One reason why larger polyremes did not take the Mediterranean world entirely by storm might be that investment in rebuilding all the ship sheds may not have been viewed with great enthusiasm.  Then there is the matter of crews ...

Yes, Anthony, the popularity of alla scaloccio did increase, apparently more or less in line with the increased popularity of using fettered convicts and prisoners of war rather than low-paid citizenry to man the galleys.  This naturally made exchanging positions in an extended voyage a somewhat unlikely exercise except for Houdini types, but it does underline your point about being the system of choice for when skilled oarsmen are at a premium.  Since it seems to have been - as far as I can deduce - Hellenistic practice to train entire crews simultaneously, taking as many men as were needed, I do not think skilled rowers were a limiting factor, and I cannot think of any references to individual crews being topped up with novices or spiced up with experts.  If you could not get the rowers, you settled for smaller ships or fewer larger ones (Philip V's navy, with its mix of polyremes and lembi, may be a case in point).

My impression is that in a near-recumbent position the legs can help the stroke (if correct this may require us to redo a bit of arithmetic about just how much power a human can provide).  One does need to raise the body a bit on recovery, but finding out how high may simply require a session on a rowing machine rather than reverting to CAD.  (I have never entirely trusted CAD since the first generation of submarines designed on it all leaked at depth - Collins, Upholder, you name it, it leaked.)

Patrick

[Patrick Waterson]

Possibly worth mentioning in the context of polyremes is this passage from Memnon's History of Heracleia:

"When he heard what had happened, Antigonus the son of Demetrius tried to cross over to Macedonia with an army and a fleet, in order to forestall Ptolemaeus; and Ptolemaeus went to confront him with Lysimachus' fleet.  In this fleet were some ships which had been sent from Heracleia, six-bankers and five-bankers and transports and one eight-banker called the lion-bearer, of extraordinary size and beauty. It had 100 rowers on each line, so there were 800 men on each side, making a total of 1,600 rowers. There were also 1,200 soldiers on the decks, and 2 steersmen.  When battle was joined, the victory went to Ptolemaeus who routed the fleet of Antigonus, with the ships from Heracleia fighting most bravely of all; and of the ships from Heracleia, the prize went to the eight-banker "lion-bearer"."

Memnon, History of Heracleia, 8.4-6 (http://www.attalus.org/translate/memnon1.html)

I have not found a Greek text for this, but the "100 rowers on each line, so there were 800 men per side" looks very much as if eight distinct oar banks are meant.

[Erpingham]

Memnon, History of Heracleia, 8.4-6 (http://www.attalus.org/translate/memnon1.html)

I have not found a Greek text for this, but the "100 rowers on each line, so there were 800 men per side" looks very much as if eight distinct oar banks are meant.

I'm not sure it will really change anything.  You know from reading Age of Galleys that some would define the eight rows as being from outside to the inside, rather than one above the other.  We would be back to interpretations of what words translated "line", "file", "bank" might mean.

[Patrick Waterson]

We can at least be certain that it does not mean eight men to an oar.  One can fit three men to an oar with some loss of efficiency, but eight???

Whatever the 'lines' were, this octoreme had eight of them.

[Nick Harbud]

Interestingly enough, it would seem (from Age of the Galley) that the majority of mediaeval galleys were one man per oar versions (albeit multiple rowers sat on the same bench), with multi-man oars being a comparatively late development and suffering from several serious problems that were never entirely resolved, notably rower inefficiency and the fact that losing one man on a bench of three disabled the entire oar (some galley captains refused to sail without a spare man for each multi-man oar).  Given Morrison's estimate that one oarsman per oar is 90% efficient and two oarsmen per oar works out at 67% efficient, one can see why few mediaeval designers ever went beyond three men per oar.

WRT efficiency of oarsmen, this is one of the subjects that was explored during the Olympias sea trials.  In particular the effect of lowest bank of oarsmen (or more commonly oarswomen due to it being the most cramped position and, hence, requiring the smallest rowers) and their contribution to the overall performance of the craft were explicitly examined.  FWIW, it was found that the lowest bank added the least to the craft's overall speed, although this was largely attributed to the lack of room that prevented these rowers from developing their full stroke.  (Incidentally, Morrison reckons the space problem was because they selected the wrong-sized cubit when designing their reconstruction.)

I hope everyone includes the Olympias trials monographs in their list of required reading on this subject.  They really are enlightening and include many practical topics such as speed, turning circle, crew training, etc.

With regard to the discussion on sources, the one name I seem to miss is Morrison's collaborator Coates, who was a naval architect.  Coates made the point that whilst many of the ancient texts could be interpreted in a number of different ways, from a practical engineering standpoint there is generally only one sensible solution.

Hope this helps,
Nick

[Patrick Waterson]

I hope everyone includes the Olympias trials monographs in their list of required reading on this subject.  They really are enlightening and include many practical topics such as speed, turning circle, crew training, etc.

With regard to the discussion on sources, the one name I seem to miss is Morrison's collaborator Coates, who was a naval architect.  Coates made the point that whilst many of the ancient texts could be interpreted in a number of different ways, from a practical engineering standpoint there is generally only one sensible solution.

Hope this helps,
Nick

Thanks, Nick.

The documentation relating to the Olympias' trials is held by the Trireme Trust and requires application for access http://tinyurl.com/k46frk4 but some of the essentials can be found here: http://www.soue.org.uk/souenews/issue5/jenkinlect.html

The late John Coates was indeed a career naval architect (he led the design of the County-class guided missile destroyers and eventually became Chief Naval Architect) and his dedication to the Olympias project resolved once and for all the question of whether a trireme had three banks of oars as indicated by sources and proposed by Morrison and furthermore how the design could be made to function.

Olympias was of necessity a bit of a botched job, as is not infrequent when pioneering uncharted territory.  It did however establish the practicability of trireme design, construction and operation, even if questionable cubits and bronze nails in place of effective measurements and wooden dowels put something of a crimp on its handling and performance.  We can regard it as a benchmark for the lowest level of trireme capability rather than a true indicator of trireme performance, proof of concept rather than a paragon of replication, but also as a huge advance in our knowledge.

Somehow I feel we are unlikely to get a skilled and dedicated team willing and able to recreate the Leophorus ('lion-bearer').  With 100 men per 'line', which I understand to be an oar bank, this octoreme would be three times as long as a trireme - and finding, let alone training, 1,600 oarspersons would tax the resources of the modern world.

Still, in a world where individuals are prepared to build a new Titanic, there is hope.

[Nick Harbud]

The Oxbow monograph on the Olympias trials is available on Amazon, although the price is slightly eye-watering.  (I found my copy in a second hand bookshop for a more reasonable £9.)

http://www.amazon.co.uk/The-Trireme-Project-Operational-Archaeology/dp/0946897581/ref=sr_1_7?ie=UTF8&qid=1370604832&sr=8-7&keywords=trireme

One wonders whether there has been any attempt to apply the interpretation/design principles used on Olympias to other ancient vessels?

[Patrick Waterson]

A worthy thought.  I am not sure they are well enough studied and understood: Coates himself wanted another try to iron out all the mistakes that cropped up the first time round, but unfortunately he had no successor and the concerted effort of design and study seems to have dissipated rather than progressed - at least that is my impression.

Does anyone know if any serious attempts have been made to apply the methodology to other classical ship types?  Ancient Egyptian ships have been analysed, especially since Thor Heyerdahl's Ra, but apart from the Phoenicia Project http://www.phoenicia.org.uk/ I know of no other attempts to recreate realistic designs of viable period vessels.

One problem is that while information about triremes was sketchy, information on the larger types is practically nonexistent.  Olympias had certain definite parameters: shipyard size, crew numbers, names for the oarsmen on the various banks, reliefs and representations giving an impression of layout and construction.  Because we can only guess at the layout and configuration of the multi-bank ships, my impression is that nobody has felt emboldened to try, although one could extend the parameters of the Olympias design as a starting-point to see at what point the equations begin to part company with feasibility and then think of ways to bring them back into the realm of the viable (the hypozomata was quite a welcome surprise, halving as it did the hogging forces on the ship).

The real problem may be that such theoretical development requires a highly skilled naval architect with time on his hands and whose findings are accepted as law if matters are not to dissolve in a crossfire of opposing entrenched preconceptions, and even then can progress only a short way before requiring the decisive test of being put into practice. 

[Gibor]

I realise its been some years, but wanted to point out there is a chaper on the Olimpias in a book "Sailing into the Past Replica Ships and Seamanship"