MarineLives is a collaborative volunteer driven project. The project started as a spinoff from a National Archives hackathon in early 2012. We are dedicated to the collaborative transcription, linkage and enrichment of primary manuscripts from the English High Court of Admiralty, 1650-1669 (with some excursions into data from the 1630s and 1640s).
Currently, we have over 10,500 images available (29 GB) and 11,400 pages and five million words of full text transcriptions on the MarineLives wiki
Our C17th ship quantitative database
We are building a database of C17th ships from High Court of Admiralty records. Copies of this database are available on request and without charge. We welcome all comments, improvement suggestions and new primary data for our database. The charts below are drawn from these data.
The latest edition of our database is version 8.9.4, issued on 04/09/2017. It contains quantitative and textual data on 886 ships, with full lists of sources.
Our volunteers make the MarineLives project special. Do please contact us if you would like to discuss volunteering, or if you have ideas to improve our wiki.
We would like to recognise and thank all those who have contributed to our project (in alphabetical order), whether as volunteer transcribers, annotators, commentators, advisors, interviewees, or PhD Forum participants.
Dr Aquiles Alencar-Brayner
Dr Roberta Anderson
Dr Gary Baker
Dr Richard Blakemore
Dr James Brown
Dr Andy Burn
Michelle María Early Capistrán
Dr Justin Colson
Dr John Davies
Dr Stuart Dunn
Professor Kai Eckert
Dr Charlene Eska
Dr Janet Few
Dr Anthony Firth
Dr Ian Friel
Professor Cheryl Fury
Dr Perry Gauci
Jamie LH Goodall
Dr Adam Grimshaw
Dr Liam Haydon
Dr Helmer Helmers
Dr Philip Hnatkovich
Rachel E. Holmes
Dr Jenni Hyde
Dr Patricia Keller
Sara J Kerr
Dr Craig Lambert
Dr Alan Marshall
Professor Steve Murdoch
Dr Shavana Musa
Dr Katherine Parker
Dr Cathryn Pearce
Professor Simone Paolo Ponzetto
Dr Benjamin Redding
Dr Mia Ridge
Dr Gavin Robinson
Dr Edmond Smith
Dr Craig Spence
Dr Chad Stolper
Dr William Tullett
Dr Brodie Waddell
Ad van der Zee
Dr Kathrin Zickermann
Dr Suze Zijlstra
and the ever helpful but anonymous @_mapnut
Summer challenge, 2017: How to make money in C17th commercial shipping?
This summer the MarineLives project team is looking at the drivers of profit and loss in C17th commercial shipping. We will publish as we go and welcome comments, contradiction, and offers of help and data.
Please contact us if you would like to learn more about this summer's project and how you can help, or if you would more generally like to learn about the work of MarineLives volunteers.
Early results from our work
Second part of the puzzle
Does a ship of a given burthen with a higher number of guns automatically have higher crew numbers (to man the guns) and therefore lower tons of freight carried per crew member?
Our tentative answer
Michael Pearce asked us on Twitter whether more guns mean more crew, irrespective of ship burden.
Below is our tentative answer:
Our data suggest greater variation in crew per gun amongst smaller ships (up to say 200-250 tons) and for ships carrying fewer guns (say up to 10 guns per ship). Moreover, the data suggest a decline in crew per gun as ship size and gun number per ship increase.
Arguably, there is a minimum number of crew per gun of 1 per gun, and that the effective minimum is around 1.4 or 1.5 crew per gun.
Ships of 250 tons burthen and below have significantly higher crew per gun:
50-99 tons = 2.5 crew/gun, n = 6
100-149 tons = 2.2 crew/gun, n = 11
150-199 tons = 1.8 crew/gun, n = 10
200-249 tons = 2.4 crew/gun, n = 18
250-299 tons = 1.8 crew/gun, n = 10
300-349 tons = 1.6 crew/gun, n = 7
350-399 tons = 1.7 crew/gun, n = 3
400-499 tons = 1.7 crew/gun, n = 3
500 tons = 1.4 crew/gun, n = 2.
Arguably, increased gun number on smaller ships and for those with relatively few guns, had no impact on crew levels because of typically high crew to gun ratios for these ships, which could simply be reduced when more guns were added, without ading more crew.
It is only when crew to gun ratios get towards 1.4 or 1.5 that additional guns, for any particular ship burden, lead to rising crew numbers and falling man per ton ratios.
Note that this dataset of n = 70 is contained within the larger dataset of n = 118, used to create a further chart, two charts below this one. For this smaller dataset we have selected only ships where we have ship burthen, crew size and number of guns. One ship has been removed with zero guns to facilitate the fitting of an exponential curve.
Tweet us or email us with your ideas.
Who can explain these data?
Our data from the 1630-1669 period from the English Admiralty Court are predominantly for English owned ships departing from London. The data suggest some increase in tons per crew member as ship burthen rises, and possibly a slight decline in guns per ship burthen as ship burthen rises
How might we explain these data?
Presented without reference to ship burthen, there appears to be an inverse relationship between tons per crew member and guns per 100 tons ship burthen.
Note that this dataset of n=71 is contained within the larger dataset of n = 118, used to create the chart below. For this smaller dataset we have selected only ships where we have ship burthen, crew size and number of guns. The smaller dataset appears to show a slightly stronger correlation between tons per crew and ship burthen.
Tweet us or email us with your ideas.
Carrying capacity of C17th commercial ships per crew member
Our data from the 1630-1669 period from the English Admiralty Court are predominantly for English owned ships departing from London. The data show a wide spread of carrying capacities per crew member for ships of a given size. Our data show only a very weak relationship between size of ship and carrying capacity per crew member. This is in contrast with Johan Söderberg (2010), who finds a strong association between the size of the ship and the ton-per-man ratio, but levelling off above ship sizes of 150 to 200 tons.
The significantly higher tons per man ratios revealed in Söderberg's data for 1692 is likely to be influenced by product mix, with outward ladings from Stockholm presumably containing high proportions of iron, tar and pitch. These products required relatively low manning levels per ton
In our own English High Court of Admiralty data, which are largely for English ships, a small number of Swedish ships have a tons to man ratio of 12.92 (n = 2) and a slightly larger number of English and Swedish ships carrying tar, pitch and/or iron from Stockholm or Gotenberg to London have a tons to man ratio of 10.74 (n = 4).
Freight charges as a function of transport distance by sea
In the 1650s freight costs for tonnage goods expressed per hundred sea miles transported vary considerably by type of goods and by degree of risk of predation. Significant premia were demanded for danger of predation during times of war, as seen in the Mediterranean and on voyages to Brazil
Transport distances have been calculated using https://sea-distances.org.
Freight charges as a function of transport distance by sea
In the 1650s freight costs for tonnage goods expressed per hundred sea miles transported vary considerably by type of goods and by degree of risk of predation. There were some transport economies according to distance travelled.
The cost of transporting goods in time of war, with much higher risk of loss, could be two or more times higher than the costs of transporting goods in times of peace for the same goods and for the same transport distances.
Transport distances have been calculated using https://sea-distances.org.
Transport costs as a proportion of final sale price
We are beginning to explore the components of transport, warehousing, port charges and customs charges as a percentage of the final sale price for a range of commodities, indentured servants and slaves in the 1650s.
The cost of lighterage per ton of sea beer from Maudlins Lane to Blackwall, both on the River Thames, was one shilling for beer worth 39 shillings per ton (2.56% of the value of the commodity) compared with an average freight rate per ton of Canary wine between 1650 and 1653 of £4.57 from the Canaries to London and an average sale price per ton of Canary wine in London in February 1650 of £38 (12.01% of the value of the commodity).
We currently have one example of the freight charges for transporting slaves from Africa across the Atlantic (from Angola to Bahia in Brazil). It is relatively unusual to see explicit freight charges for slaves, since many slaving voyages were conducted by the owners of ships on their own accounts, or by charter party per month, with monthly freight charges, as opposed to charges per slave.
Our example puts such freight charges at just over 8% of the sale price of an Angolan slave in Bahia, Brazil (3 mill 800 res to 4 mill res freight charges in 1650 versus the expected sale price per slave in Bahia in 1650 of 48 to 49 mill res).
UPDATED: How much did it cost to transport a ton of goods between ports?
C17th commercial ships could be rented by the month or by the voyage, or freight tonnage could be purchased for a voyage for a specific commodity type. We have collected freight rates for different types of commodity and for different port to port combinations and present our early analysis here. Our current dataset for tonnage based freight rates consists of seventy-two observations for a range of fine, coarse and bulk goods, with most of our observations from the period 1650 to 1666. Our data cover barley, beer, callicoes, Canary wines, coarse goods, copperis, cotton wool, cowries, currants, figs, fine goods, French wines, fruits, galls, ginger, goats wool, herrings, linen, Malaga wine, paper, pepper, salt, saltpeter, sugar, thread, and tobacco.
They include the freight rate of carrying a ton of beer by lighter in the river Thames at Blackwall in 1656 [£0.05 per ton, or an old shilling] through to the carrying of fine goods back from the East Indies to London in 1654 [£23.00 per ton]
They cover short transportation distances, such as London to Rouen and Kingsale in Ireland to London through medium distances, such as Cyprus and Scanderoone to London and Brazil to Lisbon, and long distances, such as Bantam in the East Indies to London.
Freight rates per ton were set according to distance and commodity, and reflect the different manning levels required for different types and length of voyage.
The outbreak of war had significant impact on tonnage based freight rates. For example, war between England and the United Provinces in the early 1650s, sharply pushed up freight rates on galls and cotton wool from the Eastern Mediterranean to London.
Driving the higher freight rates during times of war was the need to have higher manning levels on ships, higher mariner wages per man, and higher gun intensity per tun of ship burthen.
UPDATED: How much did it cost to transport a ton of goods between ports?
Our latest data table for mid-C17th freight rates per ton of goods transported
Our current dataset for tonnage based freight rates consists of seventy-two observations for a range of fine, coarse and bulk goods.
How much was spent on victuals per man per month on C17th commercial ships?
This is a tough question, and one we are approaching through case studies from High Court of Admiralty sources. Our best answer to date is somewhere between £1.12 and £1.44 per month per crew member, or, on a daily basis, between ca. 9d and 11.5d per day per crew member
A case study of the ship the Roebuck
(1647-1649) on a lengthy voyage in the Northern and Southern Atlantic comes up with victuals at £1.32 per month per crew member (including the master), or 10.6d
per day per crew member, compared with wages of £1.56 per month per crew member (including the master). This is a rate of
This puts victuals (that is food and drink) at 86% of wages, a proportion which is higher than suggested by examination of French naval accounts.
As context, 12d sterling per day per person was payed by Rene Louis de Cornouaille to to twelve mariners "for their dyet" for a period of five weeks, following the arrest of the ship the Julian until its release. This is £1.50 per person per month, assuming a thirty day month. The mariners were probably located in London during this time.
A lower amount of 8d sterling a day per person was assessed for "victualling" of the 90 person company of the prize ship the Julian Cavalier (also known as the Successe), which had been taken from the French by ships of the English Parliament in 1651. This is £1.00 per person per month, assuming a thirty day month.
As further context, the daily cost of a negro slave's "diet" in Angola in April 1650 is given in a High Court of Admiralty source as "ryall 40 rees or 6d sterling a day per heade", which is £0.75 per person per month, assuming a thirty day month.
Further work is clearly required, and we welcome comment from academics and non-academics on our data and analysis.
Thanks to Professor Steve Murdoch, Dr Benjamin Redding, Professor Cheryl Fury and Steve Garnett for their comments and suggestions.
What size were the ships?
The starting point of our work this summer has been to create a semi-structured database containing quantitative and qualitative data about commercial and naval ships mentioned in English High Court of Admiralty documents between 1630 and 1669. This database draws on the work of our volunteers over the last five years, with transcriptions of depositions, charter parties, and bills of sale forming the main source of data for the database.
We have ship tonnage data for 429 ships, of which the vast majority are commercial (n=407) and a small group are naval (n=22), mainly ships in the immediate service of the English Commonwealth, together with English private men of war, and a smattering of non-English naval vessels.
There are two clear peaks in the data for commercial ships - the first peak is in the 55 to 99 ton burthen category and the second peak is in the 200 to 249 ton burthen category.
Admiralty Court witnesses refer to ships of 50 and 60 tons as "small" and ships of 300 to 350 tons and above as "large". The smallest ton burthen category in our analysis (1-49 ton burthen) contains lighters, some barges and hoys, and other small river and coastal vessels.
What size were the ships over time?
We have taken a look at our growing database of commercial and naval ships from the early and mid-C17th to look for patterns over time. We still need to code by latest year of tonnage data for all the now four hundred and eight-one ships for which we have tonnage information in our database. But, for two hundred and ninety-eight of these ships, we have some basic patterns
The average burthen of these two hundred and sixteen ships is 183 tons (compared with 175 tons for the full four hundred and eighty one ships in our database).
The average burthen for ships whose latest tonnage information is in the 1630s = 176 tons (n = 54)
The average burthen for ships whose latest tonnage information is in the 1640s = 169 tons (n = 28)
The average burthen for ships whose latest tonnage information is in the 1650s = 191 tons (n = 191)
The average burthen for ships whose latest tonnage information is in the 1660s = 161 tons (n = 24)
Where were ships claiming to be English built and bought?
We have some limited information on the location of the building of ships (n = 39) and the ports where the ships were bought (n = 22). Ports of building and purchase are most frequently mentioned in cases of seizure and our data is heavily skewed towards non-English locations for building and purchase of ships. We present these information here, which are for the period 1638 to 1659. Be aware that there are a further four hundred and thirty-eight ships in our database claiming to be English for which we do not have port of building information, and most of these will have been build in England. Furthermore, there are four hundred and forty-six ships claiming to be English for which we have no information on where purchased, either from the stocks or on the secondary market for ships.
The left-hand side chart shows the location of the building of ships claimed to be English. It excludes ships claiming to be Irish, Scottish or Welsh.
The English locations for building include Norfolk, Suffolk, Essex, London and surrounds, and Suffolk. These data miss other obvious locations for ship building in England such as Newcastle, Devon and Dorset. Undoubtedly, ships contained in our High Court of Admralty database were built at these locations, but in the absence of specific confirmation they are excluded. The non-English locations for building are dominated by the United Provinces (Holland and Zeeland) and the Spanish Netherlands (Flanders).
The right-hand side chart shows the foreign ports where a number of ships claiming to be English were purchased. The data are dominated by the United Provinces, Dunkirk and Norway. These data under-represent the number of ships claiming to be English purchased from Norwegian owners, since a number of Norwegian owned ships were sailed to London and sold in the port of London, rather than in Norwegian ports.
Where were the home ports for ships in our High Court of Admiralty database?
Our MarineLives wiki contains ships from many different home ports and countries for the period 1620-1669. We have home port and quantitative information for perhaps ten percent of all ships mentioned in the wiki. We have analysed the ships for which we have quantitative data by home port.
One hundred eighty four of the ships are English, from thirteen ports; sixty-one are from the United Provinces (Amsterdam, Flushing, Rotterdam, Middleburg, Tervier, Horne, Delft); forty-two are German, from the free Hansa cities of Hamburg and Lubeck; ten are from Sweden and its dominions (Gotenburg, Masterland, Stettin, Stralsund); six are from France (Saint Malo, Brest); and three are from the Spanish Netherlands (Ostend).
What size were the smaller commercial ships?
High Court of Admiralty witnesses sometimes describe commercial vessels using specialised vocabulary. For example, hoys, lighters, and ketches. We have analysed these data to explore ship size by specialised type.
We have ship tonnage data for forty commercial vessels classified by vessel type. Clearly the sample sizes are very small for many of these vessel types.
We would be interested in our readers comments on these data.
Are the averages and ranges in the same ballpark as data in the hands of our readers, both from the C17th and earlier and later periods?
What can you tell us about the use to which these different types of commercial vessel were put?
Riverine versus coastal versus longer distance use? Cargo types? Crew and gun levels? Rental rates?
How much did it cost to rent a ship by the month?
C17th commercial ships could be rented by the month or by the voyage, or freight tonnage could be purchased for a voyage for a specific commodity type. Monthly rental rates can be found in notarised charter parties, which were submitted as schedules in support of High Court of Admiralty cases. Alternatively, monthly rental rates are sometimes recited by witnesses in their Court depositions.
Our current dataset for monthly rentals consists of forty-nine ships.
Twenty-seven of these are rental rates for hull plus apparel, tackle, furniture and ordinance, but excluding provisions and wages, which were to be paid directly by the renting agent.
Fourteen are rental rates for hull plus apparel, tackle, furniture and ordinance and including provisions and wages, which were to be paid by the ship owner and recovered through the monthly rental. We know the monthly rental rates for three of these fourteen also on the basis of excluding provisions and wages.
Finally, we have eight rental rates for which it is unclear on what basis the rentals were contracted.
How much were ships worth?
We are at an early stage of analysing value of ships. Our current data look at the value of ships for the hull plus apparel, tackle and furniture and often also the ordinance on board the ship. We distinguish between unnotarised and notarised values, with notarised values referring to specific bills of sale and unnotarised values often based upon witness estimates of ship value given in court for ships seized during voyages.
Our current dataset consists of one hundred and one ships, of which seventy-eight ship values are unnotarised and twenty-three ship values are notarised. Notarised values are lower (average = £3.90 per ton of ship burthen) comparised with unnotarised vales (average = ££6.40 per ton of ship burthen). Notarised values show a significantly tighter range around the average and mean than do unnotarised values.
We are working on disambiguating our data, but believe the differences in averages, means and range are due to the unnotarised data being more mixed in nature. Specifically, unnotarised data tends to be generated from witness statements of ship value following the seizure of a ship. We have excluded witness valuations of ships where it is clear that the outward, interim or return lading of the ship has been included in the witness valuation. Similarly, we have excluded witness valuations of ships where it is clear that an outward monetary stock has been included in the valuation.
However, even with these exclusions, the valuation of ships during their voyage usually includes some portion of the provisions carried on board the ship. If a seizure is early in a planned long voyage, these provisions could amount from anything between six and twenty months.
Moreover, the valuations of ships during their voyage will include some attribution of value to the refitting and setting out of a ship prior to a voyage, which can vary in the case of the hull from repairs to full graving and caulking or even resheathing, and in the case of the apparel, tackle and furniture, can include totally new provision of sails, rigging, blocks and other materials. Witnesses appear to make some allowance for the wear and tear of a ship on a long voyage and sometimes comment on this when giving their unnotarised valuations in their depositions in the High Court of Admiralty.
We plan to add ship inventories to our database, sourced from High Court of Admiralty appraisements of seized ships. These inventories will provide detailed breakdowns of the value of the physical components of ships in this period.
How old were the ships?
The size of our dataset for the analysis of ship age is one hundred and fifteen; of which thirteen datapoints simply state "new" or "old"; a further twenty-four datapoints are minimum ages, with the possibility that the ships were older; and seventy-eight datapoints give ship age accurate to the year. For these latter seventy-nine datapoints we have ship burthen data in tonnage for thirty-seven of them, which we display below in graphical form
The average age of ships for which we have age and tonnage data is 6.1 years (n=37), whereas the average age of ships for which we have age data accurate to the year for which we have no tonnage data is 7.1 years (n=41).
Dr Ian Friel has shared with us a summary of data from his unpublished survey of High Court of Admiralty inventory documents from the 1580s. His data are for a period forty to eighty years earlier than our own High Court of Admiralty data. Ian's survey found ages for thirty-nine ships, with an average age of nearly fifteen years and twenty-nine of them of ten years or more in age.
Comparison of textual and numerical data for 1630-1670, with the bulk of the data from the 1650s, suggest Admiralty Court witnesses regarded ships aged between zero and five years as "new" and ships of fourteen years and above as "old".
How many tons of shipping could one crew member support?
The size of our dataset for the analysis of tons of ship burthen per crew member is one hundred and sixteen. We have crew size data for one hundred and seventy-two ships, but lack ship burthen data in tons for fifty-six of these.
The average crew size for the larger dataset is 47.3, including six exceptionally manned men of war with 275 or more men per ship (n = 172). The average crew size for the smaller dataset, where we have crew number and ship tonnage is 36.5 (n=116).
How many guns were enough to defend a ship?
We have addressed this question by looking at the number of guns mounted on commercial and naval ships per hundred tons of ship burthen. The size of the relevant dataset is one hundred and twenty-nine ships, of which one hundred and six are commercial and twenty-three are naval.
There is a huge range in gun intensity per hundred tons of ship burthen amongst commercial ships, with some, particularly the small commercial ships, carrying no guns or just lightly armed, and other commercial ships as heavily armed as naval ships.
We are in the process of analysing these commercial data by geography and by commodity as well as by year to look for patterns within the commercial data.
Amongst the naval ships (a category which includes both ships in the immediate service of a state as well as private men of war under commissions from a state), there is a clear pattern for smaller ships to be particularly heavily gunned.
Digging into the data on how many guns were enough to defend a ship
The difference in gun intensity amongst commercial ships is likely to be driven by the relative value of cargo carried per ton of ship burthen and the level of predation on commercial shipping in the geographies in which ships traded.
Commercial ships carrying salt had few if any guns, in contrast to ships carrying Canary wines or cotton wool as their main return cargos.
This is likely to be driven by the low manning levels on salt ships per ton of burthen. Low manning levels meant that there were fewer people available to man guns.
We are looking at relative freight rates for salt, Canary wines and cotton wool, and at sale prices for different commodities, to see if these also drove gun levels.
Coal ships are also likely to have had few if any guns. However, most of the coal ship cases in High Court of Admiralty data concern collisions, resulting in court cases which do not ask about guns. Whereas, most of the salt ship cases in the High Court of Admiralty data concern seizures, and elicit Court cases in which gun intensity is relevant and asked about.
As we dig further into the general commercial category, we should be able to allocate a good portion of these to specific commodity groups and thus be able to improve our analysis of the drivers of guns mounted on commercial ships
How does crew size relate to gun carrying?
The size of our dataset for the analysis of crew size and gun number is sixty-nine, where we have both crew size and gun number. We have crew size data for one hundred and seventy-two ships in total and have gun number for one hundred and seventy-nine ships. Twenty-seven of the ships for which we have crew size are men of war and one hundred and forty-five are commercial ships. Forty-two of the ships for which we have gun number are men of war and one hundred and thirty-seven are commercial.
The average gun number for just men of war is 22.4 (n=2). The average gun number for just commercial ships is 12.8 (n=137). Our sample of commercial ships where we have tonnage as well as gun number (n=69) has a slightly higher average gun number than for all commercial ships, where only gun number is known.
The commercial gun number average overestimates the gun carrying propensity of commercial ships, since there is a systematic tendency not to report absence of guns from smaller vessels (vessels of thirty to sixty tons burthen). Many of these vessels, particularly those involved in coastal trade or fishing, as hoys, busses and ketches, would not have carried guns.
- ↑ HCA 13/71 f.644v; HCA 13/71 f.645r
- ↑ HCA 15/6 unfol. 111_PANA_PART_TWO_P1110620
- ↑ HCA 13/64 unfol. IMG_125_05_2784
- ↑ HCA 13/71 f.649v