DESCRIPTION AND GENERAL BACKGROUND TO ... - Springer Link

DESCRIPTION AND GENERAL BACKGROUND TO SHIPS’ LOGBOOKS AS A SOURCE OF CLIMATIC DATA R. GARCÍA-HERRERA1 , C. WILKINSON2 , F. B. KOEK3 , M. R. PRIETO4 , N. CALVO1 1 ´ and E. HERNANDEZ 1

Dpto. F´ısica de la Tierra II, Facultad de F´ısicas, Universidad Complutense de Madrid, Ciudad Universitaria, 28040 Madrid, Spain E-mail: [email protected] 2 Climatic Research Unit, University of East Anglia, Norwich NR4 7TJ, U.K. 3 Royal Netherlands Meteorological Institute (KNMI), PO Box 201, 3730 AE De Bilt, The Netherlands 4 Instituto de Argentino de Galciolog´ıa y Nivolog´ıa, CONICET, Calle A. Ruiz Leal s/n, 5500 Mendoza, Argentina

Abstract. Meteorological observations made on board ships prior to 1854 have until now been overlooked in climatological research largely as a result of the lack of rigorous abstraction and treatment criteria. However, the CLIWOC project has shown that millions of potentially valuable observations have been preserved throughout Europe in the many logbooks that were prepared by officers in the navies of the great powers. Moreover these data can be used and treated in a fashion similar to today’s ship-based instrumental observations. The most significant logbook collections are from Spain, the UK, the Netherlands and France. This paper describes the main archives where the logbooks can be found and provides an account of their contents and of the manner in which they were set out. The potential of this source is emphasised by the fact that whilst the CLIWOC project abstracted data for 300,000 days, nearly 90% of the European logbook collections have yet to be examined. This paper concludes with a discussion as to the direction of future studies in this important field.

1. Introduction Since the early days of navigation in open seas mariners have kept records of the main events on board. The documents in which these records were kept soon took the form of a logbook or journal. Their purpose was manifold: to identify new and safe routes, to transmit the accumulated experience of the most skilled officers, thus contributing to the training of young and future officers, to provide documentary evidence of the proper management of the ship, and to protect officers and ship owners from claims resulting from loss or damage to transported goods. Naval officers could use them in defence of their battle actions in the event of Court’s Martial or enquiry. It was common practice for all senior officers to keep a logbook; these included captains and lieutenants, as well as the master (navigational officer). These officers usually joined the crew as young as 10-years old and would gain notable experience before rising to a rank where logbook-keeping was mandatory. Logbooks were not, however, kept with any specifically meteorological purpose in Climatic Change (2005) 73: 13–36 DOI: 10.1007/s10584-005-6954-4

c Springer 2005

R. GARCÍA-HERRERA ET AL.

14

mind but they did include daily detail on the weather. This was needed to assist in navigation and to ensure thereby the safety of the cargo and of the lives of those on board. The need for reliable observation was therefore paramount and provides a rich legacy of meteorological information that can today be used for climatological research. Fortunately, significant collections numbering many thousands of logbooks have survived in Spain, the UK, the Netherlands and France, but in other countries, notably Portugal where a succession of fires and earthquakes might have dispersed and partly destroyed the organized original collections (Alcoforado and Nunes, personal communication) the picture is less inviting. This paper describes the main archives where those naval records have been preserved, with special emphasis on those used in the CLIWOC project and embracing the study period of 1750 to 1850. The preservation, style and form of presentation of the logbooks is reviewed with particular attention given to those aspects that deal with meteorological information. 2. Spanish Logbooks 2.1.

SPANISH ARCHIVES

The Spanish logbooks are held in two principal collections, those from the official mail ships, which are kept in the Archivo General de Indias (AGI, General Indies Archive) in Seville, and the collection kept at the Archivo del Museo Naval (AMN, Archive of the Naval Museum) in Madrid. 2.1.1. The General Archive of the Indies (AGI) The AGI holds the documents of the Spanish Administration in the American colonies and the Philippines from the time of their discovery until they became independent more than 300 years later. It was established at the end of the 18th century during the reign of Charles III to collect scattered accumulations of documents from the National Archive of Simancas, in Valladolid, which had problems with lack of space. The new collection was stored from 1785 in The ‘Lonja de Mercaderes’ (Merchants’ Meeting House) of Seville, which had previously been used as headquarters for much of the commercial activity between Spain and its colonies during the 17th and early 18th centuries. Nowadays it includes over 43,000 ‘bundles’ of documents that occupy eight kilometres of shelves, with around 80,000,000 pages in total. The organization and contents of the AGI reflects the administrative structure and management of the Spanish colonies (see Garc´ıa et al., 2001, for a more detailed description on the AGI). The source of data for the present study are those documents from the Correos (Mail) section. The mail service between Spain and the American territories was organized by the Indies Council through the system of Nav´ıos de Aviso. It maintained regular communications between mainland Spain and the principal ports in the

DESCRIPTION AND GENERAL BACKGROUND TO SHIPS’ LOGBOOKS

15

Caribbean and South America. In 1764, a modern mail service was established by Charles III, and in 1777 the Real Ordenanza del Correo Mar´ıtimo (Royal Order of the naval mail service) defined the organization, personnel, routes, and the type and size of the ships to be used (frigates). The Service worked regularly until 1793, when it was interrupted due to the French Revolutionary and Napoleonic wars, and never re-established following the independence of the American territories in the early 19th century (see Garay Unibaso, 1987, for a detailed description of the mail service). Two main routes connected La Coruña in northwestern Spain with Havana (Carrera de la Habana) and with Buenos Aires (Carrera de Buenos Aires). All captains in this service were obliged to keep a logbook and deliver a copy to the Head of the Mail Office at La Coruña on their return. These documents were preserved in the Gobernación (State) and Ultramar (Overseas) archives, jointly with all the documents of the mail service until 1864 when they were moved to the AGI. Currently they are organised as a series in the Correos section and are divided into sub-series according to routes and dates. All of the items have been described (López Gutierrez, 1996) and digitised and can be accessed electronically at the AGI. 2.1.2. The Naval Museum Archive (AMN) The AMN was created in Madrid in the 1930s as a military historical archive. It comprises four main divisions: Depósito Hidrográfico (Hydrographic Depository), Real Compañ´ıa de Guardamarinas (Royal Company of Navy Cadets), Fondos de Adquisiciones y Donaciones (Acquisition and Donation Funds), and Museo Naval (Naval Museum). The Hydrographic Depository provided most of the documents for this study. This important archive was established at the end of the 18th century to keep all ship logbooks and hydrographical information prepared by officials of the Navy during their various commissions, voyages and scientific expeditions. The collection of logbooks in the AMN comes partly from the scientific expeditions but also from the Marina Mercante y Militar (merchant and naval branches of the service). The two main routes undertaken by these naval vessels were Cádiz to Callao and Cádiz to Montevideo. Other important routes were from Cádiz to Puerto Rico, Veracruz and Havana, and finally, although not so well frequented, from Cádiz to Manila in the Philippines via the Strait of Magellan. 2.2.

THE LOGBOOKS

The Spanish logbooks usually correspond to an individual round voyage and are bound and stored as separate volumes. They include hourly astronomical observations, the estimated and observed latitude and longitude several times a day, hydrographical measurements such as soundings and currents and, most importantly, meteorological observations, which include a description of the clouds, type and amount of precipitation, air pressure, temperature, wind direction and strength, visibility and sea state as well as the course, sails set, and many other details of

16


navigation. Some of them also included maps of the coasts or islands, sketches of other ships, and even drawings of exotic birds and flowers. The bulk of the Spanish collection covers the period 1764–1800. The most important of the routine observations was the determination of the ship’s position by latitude and longitude. For many years the former was determined by using an astrolabe and a sextant to measure the altitude of the Sun. The longitude was usually estimated from a fixed reference point, such as a port of departure, then taking into account the course and distance covered each day. The estimation of longitude suffered from important uncertainties that were common to all the countries and are described in Section 3. Observations to fix the position of the ship near the coasts were much easier to calculate since they used an estimate of the distance and bearing to ports, capes or some other visible geographical object as well as charts. The nautical day onboard ship started at noon and in advance of the civil day which began at midnight, so the observations of 1 day belonged to the period from noon of that day to noon of the next day; noon being the time when the ship’s principal calculation of position was made. From noon onwards, the observations were recorded every 2 h, and a blackboard divided into several columns was used to note them. Figure 1 shows an example of such a board. The first column H, contains the hour of a day, the second and third (M and B), the nautical miles and fathoms per hour, respectively, that the ship covered at two hourly intervals. The fourth (R), the ship’s course, the fifth (V) wind direction and lastly (A), the leeway made by the ship. Leeway was the deviation from the course steered due to the pressure of

Figure 1. Board display for meteorological observation on Spanish ships.


17

the wind on the hull and sails. On the right side of the blackboard (not shown), remarks were noted such as the distance to the coast, and the observed azimuth. In the example, from 4 to 6 pm (since the day started at noon), 5 miles1 and 3 fathoms ◦ per hour were covered, with a course set at SO 20◦ O (SW 20 W), with a north wind and no leeway made. As there are no further observations of the course and wind until 8 am, and it is understood that the same course and winds were experienced in that interval. At the conclusion of each 24 h period, all of these observations were copied into the formal logbook. The ship’s course was determined by compass. Wind force was determined from the type and number of sails set and the speed of the vessel. (See Prieto et al., 2005, for a more detailed description of the wind records). The instrument used to measure the speed of the vessel was called a log and was widely employed from the 17th century onwards. It consisted of a piece of wood, the log itself, tied to a two or three hundred metre, non-elastic rope – the logline. This logline had knots to divide it into equal lengths. The log was cast into the sea, staying afloat and, in theory, remaining stationary in the water while the ship continued. The number of knots run off during a 30 s period were counted. The first knot was placed at about the same distance from the log as the length of the ship to avoid the effects of the vessel’s wake. The ship’s speed could then be calculated by reference to the time interval and the number of knots that had passed. During a 4 h watch, the log would be thrown two, three or four times, depending on any change in the wind force or the sail set. 2.3.

LOGBOOK LAYOUT

The majority of the Spanish logbooks conformed to present day A4 size or larger, portrait format and were written on both sides. The manner in which the information was organized in the logbooks varies and five categories of layout can be identified: tables, tables and events, text, mixed, and others. (a) Tables: All the information is included in table form. Normally there would be one table per page containing the date, hour, latitude and longitude, course, distance, wind direction and speed, state of the sky, and sometimes in later years when instruments were more common on board vessels, air pressure and temperature data. Figure 2 provides an example. (b) Tables and events (see an example in Figure 3): This form of logbook includes tables where some of the information mentioned above (hours, latitude and longitude, course, distance) was written, within a larger section titled ‘acaecimientos’ (remarks). Included in the latter are the most important events of the day such as storms, descriptions of the water colour (if near a coast), meetings with other ships, punishments of the crew, disease, deaths, etc. Sometimes additional meteorological observations such as sea state and cloud cover were included although the precise content might vary

18


Figure 2. A page of a typical Spanish table logbook. Reproduced by kind permission of the Archivo del Museo Naval.

from logbook to logbook with lengthier accounts spreading over more than one page. (c) Text: all the details of the journey (observations, calculations and remarks) are written down as simple, continuous text. Some of these logbooks included notes or accounts in the margins. The data were usually written on the body of the page but sometimes they include additional marginal notes. See Figure 4 for an example. (d) Mixed: this kind of logbook is very similar to the latter. They are mainly text but with some of the data (usually latitude and longitude data) arranged in table form. (e) Others: in this category are included all the logbooks which do not specifically fit in one of the other four categories but are usually a mixture of some of them. Most of them are comprised of text-based sections and tables-withevents sections. A total of 877 logbooks have been analyzed (467 from the AGI and 410 from the AMN). The AMN collection was digitized as part of the CLIWOC project, but the AGI collection had been previously digitized by the Archive itself. The


19

Figure 3. Typical Spanish table+events logbook. Reproduced by kind permission of the Archivo del Museo Naval.

most common presentational group is category b, (tables and events) accounting for 40.9% of the total number of logbooks, with the ‘text’ group providing a 29.7% of the total. Category c makes up 12.5%, and category a 4.2% (tables only). A further 12.7% belong to the fifth group, which embraces a mixture of the different presentational styles. Figure 5 shows the percentages of each category. 3. British Logbooks 3.1.

BRITISH ARCHIVES

The British archives contain the most extensive collections of ships’ logbooks and journals in the world. The various collections fully represent the wide range of maritime activities engaged in by British mariners from the 17th century onwards. There are large numbers of Royal Navy logbooks written by captains, lieutenants and navigating officers and a small number kept by midshipmen, as well as an extensive collection of journals kept by officers of the English East India Company. British archives also contain (on microfilm) logbooks of the Hudson Bay Company vessels, trading in both Atlantic and Pacific waters.

20


Figure 4. Typical Spanish text logbook. Reproduced by kind permission of the Archivo del Museo Naval.

Figure 5. Distribution of the different format types in the Spanish logbooks. 1 for tables, 2 for tables + events, 3 for text, 4 mixed and 5 others (see more details in the text).


21

Logbooks in that time are more properly termed officers’ rather than ships’ logbooks as they are prepared today. All officers on board a Royal Navy vessel kept their own logbook, correctly termed a journal. These were submitted to the Admiralty in order for them to draw their pay. The captain or commander of a vessel kept his own journal, which was marginally more detailed than those of the other officers. This additional detail was usually concerned with the general management of the ship, as he was personally accountable to his superiors for the conduct of the vessel and its crew. Not all captains wrote up their own logbooks; depending on their rank and wealth some employed a clerk or secretary and merely signed the book. For the East India Company logbooks there is less choice for any one voyage. Journals were usually kept by the captain, who was also the chief navigating officer, by the first or chief mate and often the second mate. Although ‘duplicate’ logbooks can be found for East India Company voyages, generally only one, that of either captain or chief mate, has survived. It is particularly useful to note that the effective cataloguing of ship journals has divided them into three distinct collections. The Public Record Office holds captains’ and masters’ journals. The British Library holds East India Company journals and the bulk of the National Maritime Museum’s collection consists of lieutenants’ logbooks. 3.1.1. National Maritime Museum The Museum is located at Greenwich in southeast London. The 5,205 volumes of lieutenants’ logbooks were transferred to the Museum from the Admiralty in 1938. Until 1807, Royal Navy lieutenants were required to keep a logbook for each voyage or ‘tour of duty’ and submit it to the Admiralty before receiving their pay. The volumes, catalogued under #ADM/L contain anything from 6 to 20 individual logbooks. As each ship usually carried more than one lieutenant there is frequent duplication of material. 3.1.2. British Library At the British Library, located in Central London, there are some 50 logbooks scattered among the Additional Manuscripts collection. However, most of the logbooks and journals are held in the Oriental and India Office collection and were written by officers of the East India Company. There are 3,822 journals, most of these catalogued under L/MAR/B. They are bound chronologically under the name of the ship, each volume containing three to five logs and each logbook consisting of both the outward and return leg of a voyage. Most of the logbooks cover the 18th and early 19th centuries up until 1834 when the Company ceased operations. 3.1.3. The National Archive The National Archive, formerly know as the Public Record Office is located at Kew in southwest London. It is the largest and most important archive in Britain containing official records of government from the earliest times. It holds an extensive

22


collection of logbooks and journals written by admirals, captains and masters. The Admirals’ journals are generally less useful than the logbooks of more junior officers since they do not usually contain regular entries for the ship’s position or records of wind force and direction. Many are purely narrative and are devoid of the ordered structure of formal logbooks. 3.2.

FORMAT AND EVOLUTION OF

BRITISH LOGBOOKS

The format of British logbooks was broadly consistent but differed from one service to another. Royal Navy logbooks, until the early 19th century, usually recorded the basic information of position, wind direction, wind strength and visibility at noon as well as other non-meteorological items. Additional wind directions and comments on precipitation, sea state etc. were usually listed as sub entries prefixed by either pm or am. As previously mentioned, it should be noted that the nautical day on board ship changed at noon and was 12 h ahead of the civil day, ‘pm’ therefore preceding ‘am’ in the written record (Harries, 1928). English East India Company logbooks would record hourly information, particularly course, speed and wind direction, with the ship’s position noted, as with naval logbooks, at noon. It was not until between 1805 and 1808 that Royal Navy logbooks began to record similar hourly or two-hourly information. This coincided with an Admiralty instruction to keep logbooks according to the civil day instead of the nautical day. The English East India Company, in common with other merchant ships’ logbooks, continued to use the nautical day. 3.2.1. Early Naval Logs The early naval logs followed a two (facing page) format and have been generally described in Oliver and Kington (1970). Figure 6 represents typical entries reconstructed from an actual logbook. On the left hand page of the logbook (Figure 6a) the information is recorded in columns. Column 1 gives the day name, column 2 the number day, the wind directions during the day is written in 3, the course in 4, the distance in nautical miles sailed in that day in 5, columns 6 and 7 gives the latitude and longitude after correction and adjustment and finally the bearings to the last visible mark are noted in 8. The information recorded on that page describes the vessel’s position at noon, although the wind column often recorded all of the observed wind directions for the previous 24 h. The last entry in these daily lists was the wind direction immediately before or at noon. On the right hand page (Figure 6b) wind strength and general weather information such a rain, snow, fog, thunder etc. as well as any other remarkable incidents were recorded. The day was further divided into three parts, the first part from noon until 8 pm, the second from 8 pm to 4 am and the third from 4 am to noon. The state of the weather was either recorded at these three times or as in the example, twice daily, firstly for the period after noon (pm) and again after midnight (am). Winds were recorded at least once a day at noon and


23

Figure 6. Typical 18th century British naval logbook (see text for further details). Reproduced by kind permission of the National Maritime Museum, Greenwich. (Continued on next page)

24


Figure 6. (Continued )


25

frequently more often. Their direction was described as the point from which they originated, using 32-point compass. The noon course noted in the logbook was not necessarily the course being steered by the vessel at that time but was the overall course made between the current position at noon, and the noon position of the previous day. The notation used was based on a 360◦ circle but treated in quadrants. In the example given, the course S54W (or 54◦ west of south) corresponds to 234◦ on a modern compass. The course made good was taken from a chart and this particular direction was with reference to the geographic pole or true north. The distance sailed in the previous day was recorded in nautical miles and, in common with the course, referred to the distance between two consecutive noon positions. Latitude was the ship’s position either north or south of the equator. It was determined by an observation of the altitude of the noonday sun or of the Pole Star and by reference to an almanac. Clear weather would ensure a reasonably accurate observation using an octant or sextant (Hewson, 1951). In the example (Figure 6) the vessel is just over 34 degrees south of the equator near the latitude of the Cape of Good Hope. Longitude was the ship’s position either east or west of a fixed point or a meridian. In the example the longitude 14.58E is with reference to the Greenwich meridian, placing the vessel to the west of the Cape of Good Hope. Before the adoption of Greenwich as the prime meridian in the late 19th century, London was frequently used as the ‘zero’ meridian by British ships. This was the line that passed through St. Paul’s Cathedral, about 30 s west of the present Greenwich meridian (Hewson, 1951, p. 15). More commonly before the 1790s, and particularly before the widespread use of the marine chronometer, the longitude was noted with reference to the last major sighting of land. When leaving England on a voyage to India, a vessel might use the Lizard (5.22W) or Start Point (3.63W) as a meridian for the first leg. It would later use either Madeira or more usually one of the Canary or Cape Verde Islands as a meridian until it reached the Cape, when it would then adopt Cape Agulhas (20◦ E of Greenwich) until sighting Madagascar or one of the Comoro Islands in the Mozambique Channel. This landfall would then provide a zero meridian for the final leg of the voyage to Bombay or Madras. Therefore when sailing by dead reckoning (estimating position by observed latitude, course, speed and leeway alone) the meridian used would change several times and the ship’s position, especially longitude, being corrected by sighting a known landmark such as the Cape of Good Hope or the one of the Azores islands. When reading a logbook it is essential to know which meridian is being used. In those British logbooks that used chronometers or lunar observations to determine longitude, the meridian is Greenwich. In all other instances it is usually the last major landmark sighted (frequently recorded in the Bearing/Distance column). This was the common practice but the reader should be wary of occasional deviations in recording procedures. The CLIWOC team assembled a list of over 600 meridians that were used in logbooks during the study period.

26


Bearings and distance as mentioned above were given to either the last point of land sighted or, less commonly, towards the next expected landfall. Most bearings were taken from a chart and expressed in degrees using the same notation adopted in recording the ship’s course. They were therefore, and in contrast with the wind directions, ‘true’ bearings with reference to the geographic pole (see Wheeler (2005), for a full discussion of the question of directions and compass bearings). When land was in sight, the bearing was frequently expressed as a compass point and was often a magnetic rather than a true bearing. Distances were occasionally noted in miles but more frequently in leagues, the latter being three nautical miles. There were no hard and fast rules concerning the information recorded, and although minor variations are common (and not a problem if noted) most logbooks conformed to one or other of the of the systems outlined above. 3.2.2. Early 19th Century Naval Logbooks As mentioned above, logbook layout, whilst consistent over long periods of time, did undergo change. Date keeping was one such change, and an Admiralty order of 11 October 1805 directed all officers to keep a logbook according to the civil rather than the nautical day. The presentation of information in the logbook also became more detailed with the day being considered in one or 2 h intervals. This layout was by no means new and had been used as early as the 1770s (Wheeler, 1995). The information given in these later naval logs is the same as in the older format and consists, from the meteorological point of view, principally of wind force, wind direction and weather notes. The major difference is in the more frequent attention to detailed information, with observations now scattered more liberally through the day, often being made at hourly or two hourly intervals. The pages were often prepared using printed sheets purchased at nautical stationers and suppliers at the major ports. Information was usually recorded hourly, although if the course speed and wind direction remained steady, for instance while sailing in the trades winds, information would then be recorded only at the change of watch. 3.2.3. English East India Company Logbooks Many officers of the East India Company had a seagoing experience that, in terms of sailing, covered a wider geographic area than could be claimed by the majority of Royal Navy officers. Most would have sailed as far as China at least once. They were required to keep detailed logs that recounted the state of the trade winds, the onset of the monsoon in the Indian Ocean, the strength and direction of ocean currents and the general management of the vessel (Records, 1960, p. 109). Logbooks from as early as the mid 18th century recorded navigational and weather information on an hourly basis. The example from the Melville Castle illustrated in Figure 7 is from 1787 and it was to be nearly two decades before Royal Navy logbooks contained similarly detailed accounts. During the later 18th and early 19th centuries, East India logbooks followed a consistent format with only some slight variation in the


27

Figure 7. Typical 18th century English East India Company logbook. Longitude is based on the Greenwich meridian, probably using a chronometer. Reproduced by kind permission of the National Maritime Museum, Greenwich.


28

earlier period. Towards the end of the 18th century, logbook pages were usually pre-printed forms similar to the later Royal Navy logbooks. In the example of the Melville Castle, the first column (H) gives the hour of the day. The second column marked ‘course’ gives the direction that the vessel was steering during the preceding hour. This was always a point of the compass and was a magnetic course as the steering compass, at least on English vessels, was not corrected for magnetic declination. The columns marked (K) and (F) show the speed of the vessels in knots, or nautical miles and fathoms. The latter term requires some clarification; by present day conventions this unit is six feet, and usually refers to water depth, but in the case of logbooks there are only eight fathoms to the nautical mile, giving it a measure of 253 yards, i.e. it is a nautical equivalent to the land-based ‘furlong’. The section ‘Winds etc.’ recorded wind direction and any other ‘significant’ weather observations such as squalls, cloud, haze, fog, rain or lightning. The last and largest section was for general remarks on shipboard activities, accidents, deaths, punishment given, sightings of land or other vessels, descriptions of damage to the ship, and so forth. See later chapter for examples. The remarks section also provided additional space for navigational observations such as magnetic declination. The more meticulous logbook keepers might also describe any sea swell as well as the direction and strength of a current. Some later logbook keepers would record temperature and barometric pressure. These were usually taken from instruments kept in the captain’s cabin. 4. Dutch Logbooks The Dutch contribution to the CLIWOC database is based on two different types of logbooks. The first type is the extract logbook (Koek and Können, 2005). Several copies of that type are preserved at KNMI. The second type is the regular ship logbook. These original ship logbooks are used in many variations, and are kept in a number of national archives and museums. All data from the 18th century are based on the latter type. 4.1.

DUTCH ARCHIVES

4.1.1. Royal Netherlands Meteorological Institute (KNMI) Most of the historical documents and observations from KNMI, being the official meteorological institute in The Netherlands, are stored in Het Utrechts Archief (the Archive of Utrecht). They are kept there under governmental regulations. Because KNMI was founded in 1854 the KNMI material in Het Utrechts Archief also starts from that year. However, KNMI still keeps a number of historical documents in its own archive, essentially for practical reasons. Among the latter documents can be found a number of so-called extract logbooks. In the 20 extract logbooks that fell in the CLIWOC period, the condensed meteorological data for 273 voyages are preserved. The 60,103 records from this source make up no less than 68% of the


29

Dutch 19th century observations (including 57% of the wind observations). The original logbooks, on which the extracts are based, are supposed to be lost. 4.1.2. The National Archive The main provider of regular Dutch ship logbooks is the Nationaal Archief (the National Archive of The Netherlands), which is the country’s largest public archive. The Nationaal Archief is based in The Hague, where almost 1000 years of Dutch history are stored in 93 kilometres of shelves of documents, maps, drawings and photographs. It also embraces one of the largest collections of material from the Dutch East Indian Company (VOC), but ship logbooks from other companies and from the Dutch navy are also stored there. The Nationaal Archief provided notable assistance by copying many logbooks onto film, while the films were transferred to digital images by the Netherlands Institute for Scientific Information Services (NIWI). The Dutch CLIWOC team digitized 11,673 records directly from the originals, and 35,523 records from 11,628 digital images of the originals. In total 266 Nationaal Archief logbooks were used in the CLIWOC project. 4.1.3. The Zeelands Archive (Het Zeeuws Archief) The second-largest provider of regular logbooks was the Zeeuws Archief (Zeeland Archives). The Zeeuws Archief is the principal archive for Zeeland, the most southwesterly province of The Netherlands, and is regarded as the ‘treasure house’ of its history. The Zeeuws Archief was formed in 2000 by a merger of the former State Archives in the province of Zeeland and of the Municipal Archives of Middelburg and Veere. The Zeeuws Archief has its seat in a new and well-equipped building in the center of the provincial capital of Middelburg. From this archive the Dutch CLIWOC team used only filmed material (60 logbooks; 1,830 images; 10,250 records). The films were directly available from the archive and consisted of ship logbooks of the Middelburgsche Commercie Companjie (Middelburg’s Commercial Company), which sailed almost regularly from Zeeland to the West African coast via Surinam and the Caribbean, and back to Zeeland. 4.1.4. Other Sources Apart from these large collections, ship logbooks were used from the following 10 organizations: Netherlands Maritime Museum Amsterdam, Netherlands Institute for Scientific Information Services (Amsterdam), Maritime Museum Rotterdam, Northern Maritime Museum (Groningen), Municipal Archives of Amsterdam, Schiedam and Dordrecht, The Archive of Utrecht National Museum of Natural History (Leiden) and Gothenburg University Library. 4.2.

TYPE OF LOGBOOKS

The main source of the Dutch CLIWOC data, the extract logbooks, represent summarized ship logbooks but with mainly meteorological information. In the 1860s

30


KNMI, being aware of the importance of gathering sea-based climatological information, carried out a project to extract the meteorological information from logbooks dating back to 1826. For this purpose former captains and officers were recruited. The meteorological information was transferred to the extract logbooks with a standard tabular layout. This layout was originally designed by Maury and Jansen and adopted by the Brussels Conference in 1853 (Quetelet, 1854) as the international standard. Being paid per record, the extractors developed a shorthand method of noting the weather information with many self-invented symbols. The meaning of almost all of them was established as part of CLIWOC project (Koek and Können, 2005). Several ship logbooks were transcribed into each single extract logbook. Most of the latter contain 6 observations per day. However for CLIWOC only the noon observation was digitized, being the one that also included the ships’ position. In several occasions, the original logbooks, from which the extract logbooks originate, were available. A comparison between the extract logbooks was compared and the original logbooks and were found to be sufficiently accurate. To be safe, we entered both versions into the database, flagging the extract logbooks as duplicate records. Among the regular ship logbooks several types were encountered. One type is the ‘open form’, i.e. no particular format was prescribed. The logbook keeper simply noted sequentially, largely on a daily basis, what happened during the voyage. This may include anything, from weather to punishments, exotic animals that were seen, ships’ positions, etc. Generally however, a daily report started with a description of wind and weather. Positional information, often hidden in the text, referred to noon positions, sometimes accompanied by dead reckoning calculations and estimations of the course and distance sailed during the previous day (so-called ‘course and distance made good’). In cases where the ship was sailing in coastal waters, the positional information was given in the form of bearings and distances to visible landmarks. The second type of logbook is of pre-printed tabular format. On VOC ships, this type of logbook was already in use at the end of the 17th century. In the CLIWOC period, the VOC ships used both the tabular and open formats, while the other Dutch companies and the navy used the open format throughout during the 18th century. The format in the tabular VOC ship logbooks remained more or less the same. The headings of the columns are as follows: on the left page it had the columns Month, Day, Remarks/encounters/events, while on the facing page the columns were for: Day, Course, Miles, Dead Reckoning (DR) latitude, DR longitude, Established latitude, Established longitude, Needle (compass), Winds and weather. Figure 8 shows an example of this kind of logbooks. Around 1800, the Dutch navy and merchant fleet changed from the open format to a tabular format with a standard design (see Figure 9). The headings of the columns printed on the left page are: Watch, Glasses (time), ‘Course made good’, Distance in miles, Wind direction (compass points), Wind force/weather/sky, Thermometer reading, Barometer reading, Compass orientation, Azimuth, Water at the

31

Figure 8. Typical tabular VOC ship logbook.


Figure 9. Typical 19th century Dutch logbook.

32 R. GARCÍA-HERRERA ET AL.


33

pump, while on the facing page the following column headings are printed: Notes, Remarks/manoeuvres/discoveries/events/etc. This practice persisted until the end of the period analyzed by CLIWOC (1854), the year that the international standardization of logbooks (Quetelet, 1854) was introduced on the Dutch ships.

5. French Logbooks Around 85% of the French logbooks are kept at the National Archives (Archives Nationales). They are organised in two main collections within the Marine Funds (Fonds de la Marine): those written before 1870 can be found in the Centre d’Accueil et de Recherche des Archives Nationales (CARAN) at Paris. Those dated after 1870 are kept in the Archives Centrales de la Marine of the Navy Historical service at the Chateau de Vicennes. The rest of the available French logbooks can be found at the Centre des Archives d’Outre Mer (CAOM), at Aix-en-Provence, at the National Maritime Museums, with sites at Paris, Brest and Rochefort and at different local archives at ports such as Cherbourg, Brest, Lorient, Rochefort and Toulon. In CARAN there are around 3,000 references to logbooks corresponding to the period 1594–1870. They are derived from Navy, private and West Indies Company ships and cover the routes from mainland France to the colonies in North America, the West Indies and Africa and across the Mediterranean. A more detailed description can be found at Bougin (1963). CLIWOC used French logbooks from the Archives Nationales, Section Marine, Service Hydrographique, “Marine Subserie 4 JJ 7–26: Journaux de bord”. A sample of 149 logbooks has been digitized (99 from the former and 50 from the latter). Seventeen microfilm rolls were selected to provide a comprehensive data sample. Each roll contained between 2 and 22 logbook copies. The ships’ destinations covered Canada and the Antilles Islands for a period embracing the second half of the 18th century and the early decades of the 19th century. One of the rolls contained four logbooks describing trips from France to South American such as Rio de Janeiro. These logbooks were microfilmed in 1974 for the Library of Congress, The University of Southwestern Louisiana, Loyola University of New Orleans, Memphis State University and the Mississippi Department of Archives and History. The microfilm of the section of French archival series devoted to the Louisiana region (here defined in its widest sense) was made by the Louisiana Colonial Records Project under the comprehensive terms of the Library of Congress Wilbur Fund. After 1780 navigation was based on the sounder principles of astronomical observations (by the method developed by the English astronomer Neville Meskelyne and known as ‘lunar distances’) and by use of the newly-invented marine chronometer. According to the instructions of the La Pérouse Scientific Expedition (1785–88), all marine clocks had to be verified, and angular distances of celestial objects taken to compare with the longitudes obtained using the chronometer. The

34


compass declination and inclination were also noted every day and the keeping of duplicate records was commonplace (Bonnichon, 1992). The logbooks for the period between 1737 and 1800 show different layouts. They have been classified into five categories according to the fashion adopted by the captains for setting out the daily observations. They are essentially the same as those described above for the Spanish logbooks. From around 1730, but particularly during the reign of Louis XVI, the logbooks were written on printed sheets: the officers on board only had to fill the different blank spaces. The logbooks were kept at the Hydrographic Service (Bonnichon, 1992).

6. Discussion This papers shows that, not surprisingly, the logbooks from the different nations reveal a great degree of homogeneity in their meteorological content and organization. The main differences are found in the availability of sub-daily observations and in the methods used in the determination of longitude and wind speed (Prieto et al., 2005, Wheeler and Wilkinson, 2004, Koek and Können, 2005). This general degree of homogeneity encouraged the setting up of common databases in which were entered daily noon observations of wind speed and direction and position. This aspect of the project is described more fully in Garc´ıa-Herrera et al. (2005). Despite the notable efforts directed towards the more mechanical task of data abstraction, limitations of time and funding left a large number of logbooks unexplored. It is to be noted that although the volume of abstracted data has been evenly distributed among the Spanish/French, UK and Dutch collections, the number of non-abstracted logbooks varies greatly depending on the country. In Spain, almost all of the known logbooks have been abstracted and digitised under CLIWOC, with only some dispersed and marginal collections not included. In France there are around 3,000 logbooks available for the CLIWOC period, but only a 5% of these have been abstracted. The British archives contain about 120,000 individual logbooks from the age of sail. Of these approximately 6,500 (5.5%) were examined and about 1,800 used (1.5%). Those discarded were either duplicate voyages (common with lieutenants’ logbooks), or covered areas such as the Mediterranean or Caribbean not included in the CLIWOC project, this gives an ‘attrition’ rate of about 72%. If this rate is applied to the 110,000 or so logs remaining, there are still at least 31,000 individual logs that should yield similar high-quality meteorological data. In The Netherlands it is estimated that around 1,200 logbooks are currently available in the Dutch archives, of which 40% have been digitized under CLIWOC. However, even when the remaining logbooks from Dutch archives are included the Dutch sources would not be exhausted. It was regular practice that copies of logbooks from ships bound for Batavia (now Jakarta) were stored on arrival in the eastern ports. Many hundreds of these logbooks still exist in the Arsip National of


35

Republik Indonesia in Jakarta, but their poor state of preservation and the lack of documentation make them difficult to access. Furthermore, a rigorous early 19th century ‘cleaning-up’ action in the Nationaal Archief resulted in a large number of logbooks being lost. Procedures similar to that in Batavia were carried out in Cape Town and Ceylon. However, the number of logbooks that survived in these countries is unknown, but supposedly very limited. Similar collections exist in private archives, e.g. in Germany with the descendants of Von Siebold (Können et al., 2003). These additional sources should not be overlooked in any future comprehensive review of the Dutch data. In general, therefore, it is worth re-emphasising that although the CLIWOC project has been successful in abstracting nearly 300,000 days of records, many more remain in various collections in Europe and elsewhere. With this in mind, some concluding points can be made with regard to future studies in this field. Firstly, some areas, of which the Mediterranean is a good example, have not been covered by CLIWOC and remain unexplored through this source. Secondly, the density of data points within the existing CLIWOC region can be increased to provide a more detailed and reliable impression of past conditions. Finally, the record could be extended by almost a century. This is particularly the case regarding English and French logbooks some of which date back to the 1650s. This could lead to an important extension, using logbook data, of circulation index series’ such as that for the North Atlantic Oscillation (see Jones and Salmon, 2005, for an application). Note 1

A nautical mile or 1852m being the equivalent distance along a meridian of 1 min of latitude

References Bonnichon, P.: 1992, Los navegantes franceses y el descubrimiento de América (siglos XVI, XVII y XVIII). Colección Europa y América. Ed. MAPFRE, Madrid. Bourgin, G.: 1963, Inventaire des archives de la Marine. Service Hydrographique. Sous-serie 4JJ (Journaux de bord). Deposée aux Archives Nationales, Imprimerie Nationale. Paris. Garay Unibaso, F.: 1987, Correos Mar´ıtimos Españoles. Correos Mar´ıtimos Españoles a la America Española. (Yndias Occidentales). Vol. I. De 1514 a 1827, Ediciones Mensajero. Bilbao, pp. 246. Garc´ıa, R., D´ıaz, H., Herrera, R. G., Eischeid, J., Prieto, M. R., Hernández, E., Gimeno, L., Rubio, F., and Bascary, A.: 2001, ‘Atmospheric circulation changes in the tropical pacific inferred from the voyages of the manila galleons in the sixteenth–eighteenth centuries’, Bull. Am. Meteorol. Soc. 82, 2435–2455. Garc´ıa-Herrera, R., Können, G. P., Wheeler, D., Prieto, M. R., Jones, P. D., and Koek, F. B.: 2005, ‘CLIWOC: A climatological database for the world’s oceans 1750–1854’, Clim. Change, this volume. Harries, H.: 1928, ‘Nautical time’, Mariner’s Mirror 14, 364–370. Hewson, J.: 1951, A History of the Practice of Navigation, Brown, Son & Ferguson, Glasgow.

36


Jones, P. D. and Salmon, M.: 2005, ‘Preliminary reconstructions of the North Atlantic Oscillation and the Southern Oscillation index from wind strength measures taken during the CLIWOC period’, Clim. Change, this volume. Koek, F. B. and Können, G. P.: 2005, ‘Determination of wind force and present weather terms: The Dutch case’, Clim. Change, this volume. Können, G. P., Zaiki, M., Baede, A. P. M., Mikami, T., Jones, P. D., and Tsukahara, T.: 2003, ‘Pre-1872 extension of the Japanese instrumental meteorological observation series back to 1819’, J. Clim. 16, 118–131. López Gutierrez, A. J.: 1996, Archivo General de Indias. Inventario de la Sección de Correos, Ministerio de Educación y Cultura, Madrid, pp. 794. Oliver, J. and Kington, J. A.: 1970, ‘The usefulness of ships’ log-books in the synoptic analysis of past climates’, Weather 25, 520–527. Prieto, M. R., Gallego, D., Garc´ıa-Herrera, R., and Calvo, N.: 2005, ‘Deriving wind force significance from nautical reports through content analysis’, Clim. Change, this volume. Quetelet, A.: 1854, Rapport de la Conférence, tenue a` Bruxelles, sur l’invitation du gouvernement des Etats-Unis d’Amérique, a` l’effet de s’entendre sur un système uniforme d’observations météorologiques a` la mer. Annuaire de l’Observ. Roy. de Belgique. 21, 155–167. Records of Fort St. George: 1960, Public Despatches from England for the years 1752–5. Egmore. Wheeler, D.: 1995, ‘The weather vocabulary of an eighteenth century mariner: The logbooks of Nicholas Pocock’, Weather 50, 298–304, reprinted in The Marine Observer, 67, 22–28, 1997. Wheeler, D. and Wilkinson, C.: 2004, ‘From calm to storm: The origin of the Beaufort Scale’, Mariners Mirror 90, 187–201. Wheeler, D.: 2005, ‘An examination of the accuracy and consistency ships’ logbook weather observations and records’, Clim.Change, this volume. (Received 29 June 2004; in revised form 29 April 2005)