King’s College, London
This paper examines in the forty years leading up to the First World War how different navies, particularly the British and American, approached the problem of providing “free and unrestricted movement of their fleets” by replenishment at sea. Today the deployment of a naval force to distant locations – power projection – be it a single vessel or a large flotilla, is invariably accompanied by an appropriate number of logistical support vessels. The modern replenishment ship is well versed in transferring the everyday necessities, of food, ammunition and fuel to a warship sailing along side at speed.
Two hundred years ago in the age of sail there was no fuel to transfer and a warship could be provisioned for many months prior to setting sail. That changed with the advent of the steam engine for naval use. The introduction of steam-power was rapid; starting in 1821 for the Royal Navy with the Comet and by 1837 the Admiralty was operating in excess of fifty steam vessels. As early steam engines were unreliable their use was initially confined to harbour craft that did not stray far from their fuel source. By the early 1850s the steam-powered warship had emerged as the dominant maritime weapons system. Even so warships of all sizes still carried rigging and relied albeit to a diminishing degree on the wind for their main source of propulsion, steam complemented wind power with its sole use being reserved for an emergencies or to manoeuvre in action. Only during the 1870s did steam become the exclusive means of propulsion. As the 19th century progressed so did the design of steam engines, improved among other things, by better metallurgy, vertical cylinders and triple expansion engines, all leading to greater reliability and power.
With warships wholly reliant on coal, its provision and the acquisition of coaling stations became a matter of supreme strategic importance for maritime powers. Coal was also the principal factor that determined a warship’s duration at sea. As one American commentator wrote in 1913
“We are too prone to consider that the size of our fleet in battleships alone is a measure of our national strength. That is true only in part and, like all half-truths, has a lurking within it a seed of danger. A fleet of battleships is powerful only when its constant mobility is assured, when we are able to guarantee the free and unrestricted movement of that fleet to a given theater of war, and within that area after it has arrived.”
However, before examining efforts to refuel vessels under weigh a brief examine of the how vessels were coaled in harbour will be helpful. Colliers usually delivered coal to refuelling stations as a bulk cargo. As coal stored in the open steadily looses its calorific value – the warmer the climate the faster the process of deterioration – stocks were kept to a minimum.
Vessels usually coaled along side a quay, in American and British home ports of coal sacks containing two hundredweight (101 kg) of coal were either craned or wheeled on board by local labour, tipped into the coal scuttles before being transported internally to whichever bunker required filling. Overseas on stations, such as Malta , the West Indies and China coal was conveyed on board by native labour in baskets.
However, when stocks were available afloat in a collier or lighter the coaling was carried out at anchor in the shelter of a harbour with warship supplying the sacks and the men to fill them in the carrier’s hold before they were lifted onto the warship’s deck. Coaling evolutions involved nearly all the crew and were often done to the accompaniment of the ship’s band. Coaling ship was backbreaking, dirty and dangerous work.
Speed of coaling was considered an important factor in a vessel’s overall effectiveness, and to encourage efficient coaling fleet and flotilla trophies were awarded for the one that coaled most tons per hour. Speeds were published in military journals and faithfully report to their superiors by naval attaches. Speeds though varied considerable depending on the vessel’s design, the facilities and available labour. Nevertheless, overall speeds steadily improved during the period. For example, the American Naval attaché reported in 1893 the Royal Navy coaled at between twenty to seventy-five tons an hour during manoeuvres. In 1903 the C-in-C Portsmouth informed the Admiralty “coaling from one lighter each side will not exceed 130 tons per hour” and during the First World War the Grand Fleet battleships at Scapa Flow regularly exceeded 200-300 tons per hour. But as coaling speeds increased so too did warships’ appetites, for example the 8,400 horse power HMS Inflexible of 1876 had bunkerage for 1,300 tons while the 41,000 horse power HMS Inflexible of 1907 had space for over 3,000 tons (plus 375 tons of oil fuel). The larger the bunker capacity the greater the range, but space and weight allocated to fuel had to be bought at the expense of a ship’s other characteristics, such as armament, armour, habitability and speed.
A number of factors determined the different approaches to coaling taken by British and American Navies. In South Wales Britain had the best steam coal for naval purposes. South Wales also had a highly developed infrastructure for extracting and transporting the coal to the coast from where the world’s largest collier fleet could move it to any of the global network of British coaling stations. The Royal Navy were able charter private colliers as and when required without difficulty in peacetime and considered that the large privately owned merchant collier fleet would able to cater for any additional demands in wartime. Indeed so firm was the British Empire’s grip over the world’s steam coal and coaling facilities that it was an important economic weapon in itself during the First World War. Indeed the Royal Navy was unique in its ability to project its power across the globe without recourse to logistical support from others.
The United States had no such advantages. Its indigenous steam coal was located in North Virginia and Pennsylvania ; further from the coast and inferior in quality to Welsh steam coal and on the west coast there was no suitable coal for naval purposes at all. Neither did America possess a significant merchant collier fleet or chain of overseas coaling stations. These shortcomings were came to the fore during the Spanish-American war of 1898 when the USN was forced to purchase Australian coal for its fleet in Manila and employ British colliers to help deliver it, as well as help deliver American coal along the Atlantic coast. To reduce this dependence on foreign transports during the war the USN acquired an ad-hoc fleet of twenty colliers and between 1909-1914 built a fleet of ten large colliers to be operated by the navy. These were intended to transfer their cargoes to “men-of-war within the sheltered limits of a harbour.” They were not intended to accompany a fleet in touch with the enemy; their employment was to be strategic rather than tactical. In any case none were completed in time to accompany Theodore Roosevelt’s Great White Fleet’s 1907-09 circumnavigation of the globe. Consequently problems of coal supplies were to constantly plague the venture and although 90% of the coal the fleet consumed came from American sources, American colliers only delivered on 7.5% of it. Foreign colliers – British – delivered 70% with the remaining 22.5% came from shore-based sources. A confidential memorandum admitted that the circumnavigation “could have not been undertaken had we depended upon our own resources.”
During the 1905 Russo-Japanese War the Russian Navy also found that sending a large naval force across the world was fraught with difficulties. As a belligerent nation Russia was denied help from Britain under the rules of neutrality. Britain , an ally of Japan , was not inclined to bend the rules. So, laden with inferior coal in every available space the Russians relied on hired German colliers of the Hamburg-American Line and good weather to coal in whatever sheltered waters they could find.
Although Britain held all the aces in regard to coal, the RN was not blind to the advantages of coaling at sea would be to vessels on station as part of a blockading force. Sailing ships had been able to remain on station for months whereas steam-powered warships need to refuel in harbour meant they would be absent for long periods. In 1903 Lieutenant A.C. Dewar RN highlighted the inconvenience of blockading a remote port with steamships. Citing the example of a hypothetical blockade of Toulon by 12 Implacable class battleships based at Malta , some 550 miles away. Each vessel could remain on blockade duty for 10.5 days, but would be absent for coaling for 5.8 days. Indeed it was estimated that a quarter of the Union steam ships blockading Charleston during the American Civil War were absent for this reason alone.
Of course vessels had always transferred, persons, stores and, since mechanisation, coal while at sea, but these transfers were dependent on clam weather, carried out at very slow speeds and very limited in scale. In August 1870, for instance, the Royal Navy’s Channel Squadron transferred fifty tons of coal using ships’ boats to the ill-fated HMS Captain at the rate of five tons per hour. Attempts were made to lash vessels together for coaling, but this too required calm weather and slow speeds if severe damage was not to be inflicted on both vessels. Coaling was not viable with swells “sufficient to cause a roll of more than three or four degrees, or especially to make the vessels rise and fall in the sea more than one or two feet.”
In an 1883 paper to the Royal United Services Institute Lieutenant Lowry RN enumerated the conditions any satisfactory coaling at sea system would have to meet:-
Rapidity. He considered “nothing short of 20 tons per hour for an iron clad…or 40 tons if she has several ports that can be worked for coaling, or 15 tons in a small vessel ought to be considered a satisfactory solution.” Even at forty tons an hour it would take over twelve hours to restock half the bunkers of a typical 1883 Battleship (HMS Collingwood).
Safety. Clearly the danger of injury from large weights swinging about on vessels moving at sea is greater than when coaling in a port.
The coal must be kept dry. It was generally believed at the time that wet coal was more liable to spontaneously combust and would be harder to burn.
A minimum speed of five knots to be maintained during coaling.
“It may be most import not to stop the ship or fleet entirely. With a large squadron, the capacity of whose bunkers may vary greatly, one ship may require coaling much before another, and it may be inconvenient to stop the squadron, or risky to leave one ship behind.”
Any apparatus must be stowed on the collier not the warship.
A minimum of labour should be required to operate the equipment. In port all hands are used for coaling, this would not possible when a vessel is under weigh.
Cost. Any system had to affordable and durable “enough to stand considerable knocking about at sea.”
Lowry then went on to postulate that any such system might comprise of floating watertight coal carriers transferred between vessels suspended from a cable. While those present at Lowry’s paper agreed on the advantages from coaling at sea, they differed in their solutions to the problem of refuelling. The discussion that followed went along the lines of many others when new ideas are put forward – it is best to wait until what is on the horizon makes unnecessary – as advocated by Admiral Selwyn who thought the answer lay in the introduction of oil fuel, as “I am absolutely confident you are going to see the end of coal for steam at sea altogether.” Or the idea is to solve a problem of our own making and if we remove the problem there is no need for the solution, as advocated by Captain Bedford Pim who thought the solution was the reintroduction of sails, “Now we know that no ironclad is worth much under sail. Why do we not build ships that will keep the sea under sail, the same as the ships of our forefathers did, as long as provisions lasted? In conclusion the Chairman thought the best way forward was the creation of a committee to examine the question. Although Lowry’s watertight coal carriers were rejected as impractical, his general principles were widely accepted. Indeed the advantages of developing a workable system began to exercise the minds of naval officers, engineers, entrepreneurs and cranks on both side of the Atlantic . The Admiralty alone received twenty-three submissions of one form or another between 1888-1890, each involving various combinations of cables, chutes, bridges and buckets; however none was considered suitable for a trial.
The hub of the problem was maintaining the distance between the collier and warship. Chutes, bridges and buckets required the vessels to steam dangerously close abeam and were quickly rejected by the Admiralty. However in July 1898, The Times reported that a French collier using a Temperly Transporter, an equipment developed to increase the speed of coaling in port, had transferred 200 tons of coal to two warships while steaming at six knots, the procedure only halting when the collier was damaged in a collision with one of the warships.
A warship coaling from a collier using a Temperly Transporter:
Only despatching coal along a system of cables would allow vessels to maintain a safe distance, but this necessitated some way of keeping the cable taut enough to prevent the coal receiving a dunking but not so taut to risk the cable parting. Ideas continued to arrive at the Admiralty, but it not until 1901 were any serious trials carried out when, using a cable system, the battleship HMS Royal Sovereign coaled from the collier Rosari. Relying on the collier’s winches nineteen tons per hour was transferred; to improve on this stronger winches would be required. Further trials the same year involving the battleship HMS Empress of India used, instead of winches, weights suspended from the collier’s mast to keep the line taut. Neither system was developed further; probably because the Mediterranean Fleet that conducted both trials had been able to use only equipment from local stores, thereby placing little or no burden on existing budgets so any extension of the trials would have required additional funding from London .
In 1900 the journal The Engineer reported on some American trials based on an idea put forward by the naval constructor J. J. Woodward. The system, developed in conjunction with the Lidgerwood Manufacturing Company of New York , employed cables so the vessels could steam at four to eight knots and 350ft (107m) to 500ft (150m). Unable to bring his notion to fruition before the end of the American-Spanish War Woodward seems to have returned to other work. His role was then taken over by Spencer Miller, a civil engineer and able publicist who had taken an interest in the problem since 1893. In the Lidgerwood-Miller system, as it came to be known, the vessels steamed in line ahead one towing the other, which way round depended on their size and available workable deck space. A pair of shear poles were erected on the warship from which a line with a sea anchor attached extended aft of the second vessel, the drag on the anchor keeping the line taut. Seven hundred to one thousand pounds (317-453 kg) of bagged coal was slung under a quick release hook and transported along the cable to the warship, the hook being returned along the same cable, all under the power of a winch on the collier.
The Spencer Miller system under test in 1904
Although Miller tried to sell his design worldwide it was the USN that tested it first. Over a period of five days in 1899 the battleship USS Massachusetts towed the collier Marcellus some 300-400ft behind – the rougher the sea the greater the distance – and achieved an average transfer rate of twenty to twenty-two tons of coal per hour while travelling at five knots. After the trials the captain of the USS Massachusetts is reported to have said, “There was no time during the Cuban blockade of last year when this system could not have been used. I consider it great success.” However, the twenty tons per hour fell well short of the forty tons an hour and ten knots that Miller had been informed the RN considered necessary before it would accept any system into service. The RN and Rear Admiral Melville USN proposed that the operating winches be placed on the warship instead of the collier. A logical alteration for a navy that did not operate colliers; but one strongly opposed by American constructors who regarded it as an unwelcome addition to a warship’s equipment. In order to achieve at least forty tons per hour Miller put forward a number of improvements, one of which was to replace the sea anchor by a winch with a slipping drum, so the tension on the cable could be better maintained and loads of up to 1,500 lbs carried, although Miller persisted with sea anchors for smaller vessels. He also considered that to achieve maximum efficiency from the system the construction of large 10,000 tons colliers designed to using the equipment would be required. Despite reservations, in 1902 the USS Illinois was equipped with winches and machinery for operating Miller’s system making her able “to take coal from any masted vessel it may meet in any quarter of the world”. With no machinery on the collier Miller had to resort to the sea anchor. He was strangely quite about this installation, and that it was decided later to fit his apparatus onto the Navy’s colliers suggests that the USS Illinois’s system fell short of expectations. Indeed, a few years later when the new large American colliers had joined the fleet the outcome of trials involving the dreadnought USS South Carolina and the collier USS Cyclops led the Lidgerwood company to propose installing its equipment onto six colliers. However by 1913 the General Board of the Navy was of the opinion that even sixty tons an hour was insufficient to meet fleet requirements, and as new ships entering service had greater range it was not expected that they would need to refuel at sea anyway. Also the tactic of close blockade had become obsolete and colliers would be a encumbrance to any fleet in proximity to the enemy. The Board therefore did not see “any military necessity for a coaling at sea apparatus”. Lidgerwood managed to challenge the Board’s decision strongly enough so that in October 1914 the Board relented and recommended, “that three more colliers in addition to Cyclops be fitted with the apparatus,” two each for the Atlantic and Pacific Fleets. 
The RN also tested Miller’s system. Using new winches perfected by the British Temperley Company, it became know as the Temperley-Miller system. The Muriel, the collier chartered for the trials in 1901, was smaller than Miller would have liked. Nevertheless, dock trials commenced in November, and after a few glitches were ironed out sea trials began in February 1902 using the battleships from both the Channel and Mediterranean fleets. A mean transfer rate of forty-seven tons per hour was sustained for three hours, and at one point HMS Vengeance achieved a transfer rate of sixty tons an hour before a line parted. The limiting factor was the inability of the collier to feed the cableway fast enough because of limited deck space and labour. Although able to transfer up to sixty tons an hour, to sustain this rate would require larger colliers with larger crews and the coal bagged ready for transfer. Requirements that conflicted with the British system of chartering colliers, as private colliers were generally between 1,500 – 5,000 tons and carried their cargo in bulk, relying on the receiving warship to provided the sacks and labour for bagging. Nonetheless, although no further sets of equipment were acquired, the trial equipment was not discarded and retained in store at Portsmouth .
Undeterred by the lack of sales to the RN ever the optimist Miller persuaded the Russian and Italian Navies to test his equipment. A month after the outbreak of the war between Japan and Russia, the Russians were persuaded enough to order eight complete sets of the Lidgerwood-Miller system to fit onto the warships of the Russian Second Pacific Squadron. However, there appears little or no evidence of them being employed during the Squadron’s ill-fated voyage to the Far East .
Another scheme put forward to the Admiralty was that of Thames Iron Works, Blackwall, the company proposed “to construct one or more fast coaling transports to carry coal out to the distant fleet” equipped with its patented ‘Express’ equipment that would enable vessels to coal at sea at a rate of fifty tons per hour. Unlike Miller’s reciprocating system, the ‘Express’ used an endless cable that delivered coal broadside to the warship and returned the empty sacks to the collier, the line being kept taut by a counter weight suspended in the collier. However the Admiralty still regarded broadside coaling as too risky, and were concerned that if the sixty-ton counter weight broke free it would plunge through the bottom of the collier. To eliminate the Admiralty’s concerns the company modified the system, and submitted a revised version that included additional safety fixtures on the counter weight and end on coaling with the warship ahead. Nonetheless, the transfer rate of fifty tons per hour failed to impress the Admiralty, who now thought “a rate of at least 75 tons should be guaranteed with the prospect of 100 tons”. The Admiralty also considered it undesirable to pay the at least £7,000 that any trial would cost. (The Spencer Miller trials had cost £4,000 for the equipment and £1,000 for the alterations to the collier plus the charter costs.) Informed in 1903 that the Admiralty were not willing to proceed with the matter Mr. Mackrow of Thames Iron Works claimed that the land trials had achieved a transfer rate of 150 tons per hour and the company had had received inquiries from Germany, Italy, France and Japan, but he “had held off to let the Admiralty have if not the exclusive rights at least first go”. Moreover the company would pay for the fitting out of a collier if the Navy provided a warship and agreed to buy the equipment if the trials were successful. The offer was accepted provided the cost of the equipment was made known before hand and it coaled at, at least eighty tons per hour for six hours. But the trials failed to take place, as in October 1904 the Admiralty denied the company’s request to use an Admiralty coaling lighter for further tests before it paid to fit out a collier.
Designs for coaling at sea also originated from within the RN. The schemes of the early service advocates, such as Lowry, Bell and Tupper were obviously impractical, but the one put forward in February 1902 by Chief Engineer Metcalf was considered, albeit with reservations, worthy of a trial. Metcalf’s scheme envisaged the warship towing the collier with the tension on the cable being maintained by a steam ram. At very little cost the initial trials commenced on shore at Chatham in early 1903 and were considered promising enough to justify progressing to sea trials using the old warship HMS Basilisk as the collier. The original cost estimates for sea trials were soon exceeded and by the time HMS Basilisk was ready the total expenditure had risen to £2,200.
Captain Wonham a retired coaling officer who had overseen the Miller trials over saw these too and reported in November 1903 that they showed that nine-point-nine knots was the best speed to operate the equipment and fifty-four tons per hour was possible in smooth water. The equipment was even tested at night out under arc lamps, with twenty tons being delivered in twenty minutes to the battleship HMS Revenge. However, the Director of Stores questioned if collier owners would be willing to have the steam ram fitted on their ships in peacetime and if not the delay in fitting it in an emergency may involve serious risk, therefore it should be considered that the ram should be a fixture on the warship. The Basilisk’s limited coal capacity precluded any prolonged evolutions or endurance tests consequently the 5,750 ton collier Torridge was chartered, attached to the Channel Fleet and loaded with 2,000 tons of ready bagged coal ready to commence further trials in June 1905. The captain of the battleship HMS Duncan reported that the system “should be able to supply coal in ordinary weather to a ship steaming nine or ten knots, at the rate of eighty tons per hour, for which 100 men would be required on board the collier, if the coal was in bags.” One reason Metcalf’s system achieved a high delivery rate was because it employed two endless cables, one each side of the ship and both transporting coal, while tensioned by the same steam ram. (HANDOUT) According to Captain Wonham this system, utilising endless cables and divided loads, was unquestionably superior to Spencer Miller’s single line reciprocating method. However, in did increase the amount of wear on the cables making the splices liable to fail and in the event of an emergency the vessels could not be easily parted unlike Miller’s device which could release them almost immediately. This latter point Wonham considered important and should he thought be addressed by design changes, while the life of the cables could only be tested by exhaustive trials.
HMS New Zealand coaling using Metcalf method:
The trials continued into 1906 by which time it was thought that enough information had been gained. Metcalf, now a Commander, thought the trials successful enough to suggest “That the apparatus should be supplied and fitted to the warships as part and parcel of her general equipment [and] that the apparatus should be provided for fitting in the colliers.” In contrast the Director of Stores thought trials had not “resulted in a conspicuous success” and that both Metcalf’s notions impracticable because of the cost – £2,000 per set – and because of the weight – twelve tons – which “in the shape of top hamper is a very serious matter”. Moreover, the Director of Stores wanted time to evaluate reports of an improved system by a M Leue before making a decision.
To fully consider the matter a conference attended by the Secretary of the Navy, the Directors of Stores and Contracts, Assistant Director of Construction and four naval officers was held by the Admiralty on 3rd December 1906 . The committee was of the opinion that the advantages of being able to replenish at sea in time of war were considerable and it was desirable to adopt some form of apparatus to accomplish this. Although the committee were aware of the drawbacks of Metcalf’s system and suggested further trials with smaller warships, it regarded his as the best available and recommended its adoption. It was though a partial endorsement, as it considered just three or four sets sufficient, and that these should be installed onto colliers not warships. Leading to the question of which colliers? Having the Navy operate the colliers would, the committee concluded, be the most efficient, but also an unjustifiable expense in peacetime. If the gear were permanently fitted into subsidised colliers the purchase of additional sets would be necessary if the right number were to be on hand when required. The committee settled on providing the gear to colliers on extended time charter.
Even this limited endorsement was not unanimous as the Assistant Director of Naval Construction W. H. Whiting questioned the importance of coaling at sea “for a nation which has a great numerical preponderance in coaling stations and in ships, and whose ships are generally larger and carry a larger coal supply than those of the same class belonging to foreign powers.” Any coaling at sea he deemed to be too risky as it involved the transfer of part of a battleship’s crew to the collier and “coaling en voyage means a great reduction of speed for the whole fleet, if even a single ship is being coaled. Despite the Committee’s recommendation for the purchase of three or fours sets of apparatus in January 1908 the Director of Stores decided to buy only the one set. So far no evidence of its use has come to light, and the comment following the 1912 annual manoeuvres that every opportunity should be taken to practice coaling (and oiling) at sea suggests that it was not used.
If the Director of Stores’ decision to only purchase one set of Metcalf’s gear was predicated on hopes that M Leue’s system would prove superior, he was to be disappointed. Trials in July 1905 by the German Navy of Leue’s system demonstrated that the equipment, which like Metcalf’s used a ram and endless cable but also had a vessel quick release facility, fell short of the designer’s expectations of seventy to one hundred tons per hour, as a maximum of only fifty tons per hour was achieved. Moreover the tensioning apparatus at twelve point five metres long one point eight metres high and two metres wide, was too large to be fitted on a warship and consequently restricted to colliers, a serious drawback for the German navy that had hoped to be able to coal from any steamer a warship might meet.
One of Leue’s fiercest German critics was Otto Adams, who not surprising had his own system under development. Adams’ system was cheaper and smaller than Leue’s, and like Metcalf’s it used two endless cable but employed an electric motor connected to a system of pulleys to provide the tension on the cable necessary to transport, so Adams claimed, a combined total of one-hundred and twenty tons per hour. Tested in 1909 using the armoured cruiser KMS Roon, transfers of sixty-five to ninety tons per hour were achieved, and it appears that few if any sets of Adams’ apparatus beyond the trial equipment was purchased by the German navy. The USN that had also shown a keen interest in it also declined to buy Adams ’ apparatus.
Apart from the USN training exercises, the writer has seen no evidence that any of the belligerents coaled at sea during the First World War. A number of factors pushed the notion down naval planners’ wish lists. For the British, with their stranglehold on the world’s coaling stations, there was little or no need to coal vessels in transit while they were at sea. The attraction of refuelling the blockade line on station disappeared when the close blockade was replaced by the distant blockade. Indeed the vulnerability of vessels following a constant course and speed as they coaled under weigh, made them easy prey to submarine and surface attack, reinforcing the reasons that caused the abandonment the close blockade in the first place. Also, as the RN depended on private colliers fitting special naval equipment into private vessels would incur additional cost and problems of training and availability.
On the other hand, as it discovered during the Spanish-American War, the American Navy’s need to coal at sea was more compelling. But by the time the Americans had commissioned large colliers more suited to the existing technology it was apparent that the use of coal was on the wane, hence the limited issue of Miller’s equipment. And after America ’s entry into the First World War the USN in the European theatre was able to rely on British resources for coal. The French, Russians, Austro-Hungarians, Italians and German High Sea Fleet did not stray far enough during the First World War to require replenishment at sea. The warships that did need to coal at sea were the German raiders of 1914, but even they did not carry the necessary special equipment.
No navy considered that the capacity of any of the many systems for coaling at sea on offer before the First World War showed sufficient promise to warrant allocating them a significant portion of their precious budgets or reorganising their logistical agreements to suit the system’s requirements. As none of the systems’ designers managed to keep pace with the ever increasing demands of the navies the dream of deploying fleets across the globe that could rely on their intrinsic coal resources never materialised. Only when oil replaced coal was the available technology able to match the ambition, and, as predicted by Admiral Selwyn in 1883, it become possible to rapidly and efficiently refuel at sea. David Lyon & Rif Winfield, The Sail and Steam Navy List, Chatham , London : 2004. p. 148.  ibid. p. 19.  Denis Griffiths, Steam at Sea, Conway Press, London : 1997. p. 119.  C. Theo. Vogelgesang, Logistics –Its Bearing Upon the Art of War, USNIP, Vol. 39, No. 145, 1913, p. 68.  Charles Owen, Yarns from the fleet, Thrupp. Sutton: 1997. p. 54.  Memorandum Office of Naval Intelligence (ONI), Rate of Coaling English Navy, 17th April 1897 . United States National Archive (USNA), Record Group (RG) 45, Box 708 .  C-in-C Portsmouth to Admiralty, February 1903. United Kingdom National Archive (NA), ADM 1/7675.  Logs of HMS Hercules 1914-18, NA ADM 53/44166-70.  Oscar Parkes, British Battleships, Seely Service & Co. London : 1957. pp. 252 & 492.  Memorandum on the Urgent Necessity of an Adequate War Supply of Coal. 4th March 1910.USNA. RG. 80 Box 39 .  Paul Silverstone, US Warships of World War 1, Ian Allen, London : 1970. p. 206.  General Board Memorandum, ( 28th April 1908 ). USNA RG. 80, Box 114.  USNA RG. 38 Box 837 .  Memorandum, 4th March1910, USNA RG. 80, Box 39.  Lamar J. R. Cecil, “Coal for the Fleet the Had to Die”, The American Historical Review, Vol. LXIX, No. 4, July 1964, pp. 990-1005.  USNIP, Vol XXIII. 1897, p. 365.  R.S. Lowry, ‘On Coaling Ships or Squadrons on the Open Sea ’ Royal United Services Institute (RUSI) Journal 1883, p. 386.  ONI, Notes on Coaling Warships, Washington : 1899. p. 15.  Cit. op. 16.  R.A. Burt, British Battleships 1889-1904. Naval Institute Press: Annapolis , 1988. p. 26.  Memorandum by Capt. Wonham, 1905. NA. ADM 1/7827.  Coaling Vessels at Sea, USNIP. Vol. XXVI. 1900, p. 211.  Cit. op. 24.  “Coaling at Sea”, The Engineer, Vol. 89, 27th July 1900 . pp. 84-86.  Spencer Miller, Coaling of the U. S. S. Massachusetts at Sea. Transaction of the Society of Naval Architects and Marine Engineers, ( New York ) Vol. VIII, 1900. pp. 155-165.  USNA. RG. 80, Box 114.  Spencer Miller, Coaling Warships at Sea – Recent Developments. Transaction of the Society of Naval Architects and Marine Engineers, ( New York ) Vol. XII, 1904. pp. 177-199.  Cit. Op. 24.  Memorandum by Controller of the Navy, 28th November 1903 . NA. ADM1/7748 .  Correspondence, November – December 1904, NA ADM 1/7824.  NA. ADM 1/8725.  Report by Captain Wonham, 1905. NA. ADM 1/8727  Director of Stores to 4th Sea Lord, 22nd October 1906 . NA. MT 23-201.  Report of Conference on Coaling at Sea held at the Admiralty on 3rd December 1906 . NA. ADM 1/8004  ibid.  NA ADM 1/8269  Jahrbuch der Schiffbautechnischen Gesellschaft Vol. VII, 1906. p. 489, USNA RG. 38 Box 835 .  USNA. RG. 38 Box 837 .  The opening of the Panama Canal in 1913 also reduced the requirement for the USN.  Edwin P. Hoyt, The Last Cruise of the Emden , White Lion, London : 1975. p. 150.