In July, 1904, in one of the charming villas overlooking the city of Bar Harbor a meeting took place, small as to numbers but weighty with importance in the affairs of the Peary Arctic Club, for at that meeting was taken the formal step which meant the building of the Roosevelt.

Four men were present at the meeting: Morris K. Jesup, Lewis L. Delafield his counsel, Captain Charles B. Dix, and myself.

Mr. Jesup had stated some time previous, that if subscriptions to the Peary Arctic Club could be secured to the amount of $50,000, including his own generous check for not less than half that sum, he would assume responsibility for the construction of the ship and guarantee the contract, thus insuring the construction of the ship in time to go North in 1905, and giving nearly a year additional time in which to secure the additional funds necessary.

Up to this time the interest had not been particularly widespread. The amount of subscriptions was still short of $50,000, but time was pressing and the material must be ordered at once in order to give even a reasonable chance of completing the ship in time.

Personally I felt no doubt but what the total amount of money could be raised, and yet it must be admitted that the prospects were none too favourable and discussion did not seem to appreciably clear the situation.

Mr. Jesup was as deeply interested as I, and was not only willing but anxious to do everything in his power to put the matter through, but he hesitated at assuming too much responsibility because, as he frankly told me, he did not feel, much as he wished to, that he could properly assume the entire burden of the expedition.

Finally Captain Dix said that he would order the timber for the building of the Roosevelt on his own responsibility; that he believed the money would be raised, and that if it were not, he would assume whatever loss might result from his action. His statement was like a ray of sunlight both to Mr. Jesup and myself, for it brought out clearly the fact that there was something in the project which appealed irresistibly to business men of big ideas.

The next scene which I recall most distinctly was in another beautiful villa in Vermont, commanding miles and miles of beautiful country and with a regal mountain and forest domain back of it. It was just before the 1st of August, the date on which the $50,000 must be subscribed to insure the signing of the contract for the construction of the ship. The total still fell several thousands short of that amount. Mr. Colgate had already promised a generous check with an intimation that he might increase it if it were necessary.

At this meeting there were but three: Mr. Colgate, Judge Darling, Assistant Secretary of the Navy, and myself. The situation was presented to Mr. Colgate, and with characteristic promptitude and generosity his check was increased by an amount that rounded out the $50,000 and so the building of the Roosevelt became a certainty.

In approaching the general question of a ship for Arctic or Antarctic ice navigation, one thing is immediately apparent to anyone at all conversant with the matter, i. e., that she should be as small as is consistent with carrying the party, supplies, equipment, and coal for the work planned.

The reasons for this are evident. The smaller a ship is, the stronger she is, and the more easily handled.

In looking for facts to show the results of past experience in this field, it is at once discovered that practically all ice boats past and present have been built by the three countries, Scotland, the United States, and Norway, for the prosecution of the whale and seal fisheries.

In this work the Norwegians have operated in the seas about Spitzbergen, Jan Mayen, East Greenland, and Nova Zembla; the United States in Hudson Bay and Bering Sea; and the Scotch principally in the chain of waters comprising Davis Strait, Baffin Bay, Lancaster Sound and their tributaries, with a few voyages to East Greenland and Hudson Bay.

The ice conditions encountered by the Norwegians and Americans may be very broadly stated as floes and broken ice drifting in an open sea, through which the ships have to thread their way.

The ice conditions encountered by the Scotch whalers, are a nearly solid expanse of one season’s ice in Melville Bay, and when that is passed, heavy ice in narrow land-locked channels, notorious for their strong currents, the direction of which is opposed to the course of the whalers.

It has been said by one writer that the American whalers use their steam to keep out of the ice, while the Scotch use theirs to get into and through it.

Comparing existing ships of the Scotch, Norwegian, and United States whaling fleets, it is found that the following average proportions of beam to length exist:

It is seen at once that the Norwegians and Americans have not departed from the old-fashioned sailing ship model. (The average ratio in our modern Bath-built schooners is 1:4.78.)

The Scotchmen have a finer model, and since this model is a practical evolution by shrewd seamen and builders from an experience of over one hundred and twenty-five years, in a business where large financial returns were the lot of the best ship; and the seas where that experience was secured and for which that evolution was designed, are the seas to be navigated by the proposed ship, it seemed clear that the Scotch model was the one on which to base our studies.

The problem of size did not present itself in the present instance in quite the form that it did to Nansen, and the English and German Antarctic Expeditions. In these instances the size of the party and the length of time it was to be absent being determined upon, and the coal consumption of the engines fixed, it was easy to calculate the cargo to be carried which, plus the dead weight of the ship and machinery, gave at once the displacement needed.

In the present case it was regarded as practicable to determine in advance upon a size and proportion of ship which should most nearly balance and meet the various requirements, and let the difference between her displacement, and her own dead weight, go for cargo capacity, of which the greater portion would be coal.

The size fixed upon was 184 feet over all by 35 feet beam by 16 feet draft, loaded. (Load water-line 166 feet.) This gives a ship of nearly the same length, but a little greater beam than the English Antarctic ship, Discovery. Her length ratio would be 1:5.26, not quite as fine as the Scotch average, but much finer than the Norwegian or American models.

Such a ship is in the same class as the Terra Nova, Bear, Thetis, and Neptune of existing whalers, the Proteus (lost), and the exploring ship Discovery.

Length and beam having been determined, the form of hull was next to be considered. In the navigation of the particular regions contemplated by the Expedition, a light draft is preferable to a heavier one, as enabling the ship to go closer to the shore, and thus get round a barrier, or retreat close in shore from advancing heavy ice, and let it ground outside of the ship.

The element of light draft also enters into the consideration of the lifting of the ship under heavy pressure from ice-floes. The deeper a ship is in the water, the more difficult will it be for her to rise and save herself.

It has been well said that while a form of hull that would allow a ship to rise easily and readily under ice pressure is desirable, and this desirability has been recognised, no ship previous to the Fram had been built to meet that requirement.

In the Fram almost everything else was sacrificed to this requirement. Seaworthiness was sacrificed, and as the Fram’s experience in her two voyages shows, ability to make her way through ice was sacrificed.

For the purpose for which she was designed, i. e., to enter the ice and then drift with it, evading destruction from ice pressure, she was well adapted, but as the designers of the German Antarctic ship Gauss said in discussing the Fram model, she would have been even better adapted for this had she been bowl-shaped.

Contrary to popular ideas, the work which an Arctic ship has to do is not principally that of breaking up one season’s ice, as is done by harbour and river icebreakers, in Canadian and Russian waters for instance. Such conditions of level, unbroken ice of uniform thickness are found only in Melville Bay on the upward voyage, where the one season ice is encountered, and late in the season when the new ice is beginning to form. The main work of the Arctic ship is that of threading and pushing and wedging and prying her way among and between and around fragments and cakes and large floes of ice, the latter of such thickness (twenty to fifty or seventy feet) that nothing could break a passage through them. Of course, nothing can be done but squeeze a way around these. It is for this reason that the powerful Russian Ermack is not available for a Polar voyage, and why she is not treated of in this discussion. Fifty Ermacks merged in one could not break through these floes, and in squeezing around them the Ermack could not carry enough coal to take her half-way to the Pole.

To return to the hull model. In the Fram everything was sacrificed to secure certainty of lifting under pressure. In the Gauss, which is a modified Fram, while the broad beam of the Fram (thirty-six feet) was retained, greater length was given the ship to render her a better sea boat for the long voyage from Germany to the Antarctic Circle. Her ratio is 1 to 4.25 as compared with the Fram’s 1 to 3.25. The Gauss’s draft, however, is excessive (nineteen feet).

As already noted, great draft is a disadvantage in the region under consideration, and every increase in beam makes impassable leads which otherwise would be available, and greatly increases the power required and the difficulties of pushing a way through loose ice.

The English Discovery was built, as was to be expected, on the lines of the Scotch whalers, with a little broader beam. Her ratio is 1 to 5.27. Her draft is a little less than that of the whalers. She was not specially modelled to rise under pressure, but was specially constructed (as the Fram and Gauss were not) for ramming a way through opposing ice.

The model selected for the Roosevelt was intended to meet the requirements of lifting under pressure, of being short enough to handle easily, and of being able to ram a passage through heavy ice effectively and continuously.

Detailed features of these requirements are as follows:

For lifting under pressure, steel-sheathed sides, sloping bilges, flat floor to prevent heeling when lifted, flush stem and keel, raking stem, raking stern (this a new feature). For forcing a way through loose ice: sharp wedge bow, and full counter to keep ice from propeller. For ramming ice: a sharply raking stem, steel-sheathed.

From this general description, it will be understood that while the hull model contained the best features of preceding ships, it was not a departure from ordinary models, like the Fram and Gauss, but rather a modification of them to meet special requirements.

When the question of power was approached, there was a radical departure, in fact a complete reversal of previous practice in Arctic ships, and the adoption of ordinary commercial practice.

Hitherto Arctic ships have had full sail power (full-rigged bark being the favourite rig) and auxiliary engines, often of surprisingly puny power. The object of this has been economy of coal, and the consequent ability of the ship to cover long distances at slow speed, and remain away from home for a period of years.

The Roosevelt is a powerful steamer, with all the engine force she could contain, and with only moderate sail area. There is no question in my mind but that this is the correct principle upon which to build a modern Arctic ship for effective results.

The Smith Sound or “American” route is especially advantageous for this method, presenting a coasting voyage, facilities for placing coal depots, the key of the route condensed in a few hundred miles of heavy ice navigation, and the possibility of even obtaining coal in situ along the route.

The Roosevelt had engines capable of developing one thousand horse-power. They were of the inverted, compound type, driving a single eleven-foot propeller, and steam was supplied by two water-tube boilers and one Scotch boiler. Her sail plan is a light, American, three-masted schooner rig, possessing the advantage of light weight (it is to be remembered that every pound of weight saved in rigging or fitting means a pound of coal in the hold), and small surface to be forced through a head wind; yet sufficient to materially help the engines in a favouring wind, and to enable the ship to make her way home should her coal be exhausted.

As to construction: The strength of the hull must be such that it will resist the terrific pressure of the ice-floes, and keep its shape intact until the lifting of the ship bodily releases the pressure; such that if supported at each end only, or in the middle only, or thrown up on the ice and resting upon her bilge, during the paroxysms of the floes, she will not be strained or injured; and such that she can ram the ice by the hour when necessary, without injury to seams or fastenings.

It is a popular fallacy that steel is a suitable material for the construction of an Arctic ship. A steel ship, though structurally strong, is peculiarly vulnerable locally to the ragged, rock-like tongues and corners of heavy Arctic ice.

The elasticity, toughness and resiliency of thick wooden sides are essential in an Arctic ship; but the wood planking may be steel-sheathed on the outside to enable the ship more easily to slip from the grip of the ice, and the methods of composite ship building may be utilised in the interior of the vessel, to reduce weight, while at the same time increasing its structural strength, and not lessening the strength and rigidity of the interior bracing.

In the interests of strength, the frames of the Roosevelt were made treble, keel, keelson, stem and stern-post exceptionally strong; the planking is double; the deck beams, and especially the ’tween-deck beams, which are to be just below the water-line, are extra heavy, and spaced more closely than usual. Additional struts from the bilges, and strong posts from the keelson, longitudinal tie plates at the waterways and on the upper-deck beams, and transverse bulkheads, add still further to her great strength.

In the interest of lightness there is no ’tween-deck planking, no interior fittings; and the spars and rigging are made as light as possible.

The keel, false keel and keelsons are of oak, and form a rigid backbone to the ship six feet in height. The stem and rudder and propeller posts also are of massive oak timbers, the former having a depth on the ship’s axis of seven to ten feet, to take the blows when ramming ice. The frames also are of oak, placed almost close together, and each composed of three thicknesses of timber bolted together to give great strength. The planking is double, yellow pine inside and oak outside.

The sides of the ship are from twenty-four to thirty inches thick.

To keep even these heavy sides from being crushed in, they were reinforced by heavy deck beams placed unusually close together, and a lower tier of heavy beams just below the water-line forming with steel rods and inclined posts and struts to the ship’s sides and bilges, a strong truss at an interval of every four feet in the length of the ship.

The housing of the personnel of the expedition in light structures on deck, which personal experience has shown to be much the simpler and better plan than below decks, permits a stronger and more effective arrangement of these trusses than has been attained in previous ships. The interior of the bow, which is to the ship what the cestus was to the ancient gladiator, is filled in solid with timbers and iron.

The stern also, as well as the stem, is iron-plated, and the rudder post, which is the Achilles’s heel of an Arctic ship, is of unusually strong construction. The rudder is so arranged that it can be hoisted on deck out of the way of the ice if necessary. The propeller is so arranged that it can be used either as a two-bladed or a four-bladed propeller, and is made of unusual strength. Powerful deck appliances in the shape of windlass, steam capstans and winch, enable the ship to warp herself out of a dangerous place, or pull herself off the bottom should she get aground.

The whole plan and theory of the ship was, first, that her strength, her power, her weight, her carrying capacity, should all be below the main deck, and that everything above deck—houses, bulwarks, spars, sails, rigging, boats and equipment—should be as light as possible, to permit more coal in the hold; and second, that not a dollar was to be wasted on fittings or frills, everything to be for strength, power, and effectiveness.

The keel of the Roosevelt was laid October 15, 1904, in the McKay & Dix shipyard at Bucksport, Maine, and the ship was launched the 23d of March, 1905, Mrs. Peary shattering a block of ice containing a bottle of champagne against the steel-clad stem as the hull glided down the ways and christening the ship Roosevelt.

The installation of the machinery began two days later at Portland, Maine, and was practically completed in less than two months.

The official measurements of the ship are as follows: length, 184 feet; breadth, 35.5 feet; depth, 16.2 feet; gross registered tonnage, 614 tons; maximum load displacement, about 1,500 tons. The backbone of the ship, viz. keel, main keelson, stem and stern posts, as also her frames, plank sheer, the waterways, and garboard strake, are white oak. Beams, sister keelsons, deck clamps, ’tween-deck waterways, bilge strakes, ceiling, and inner course of planking, yellow pine. Outer planking, white oak, and decks, Oregon pine. Both the ceiling and outer course of white oak planking are edge-bolted from stem to stern and from plank sheer to garboard strake. The fastenings are galvanised iron bolts, going through both courses of planking and the frames, and riveting up over washers on the inside of the ceiling.

Special features of the ship are as follows:

First, in model, a pronounced raking stem and wedge-shaped bow; very sharp dead rise of floor, affording a form of side which cannot be grasped by the ice; a full run to keep the ice away from the propeller; a pronounced overhang at the stern to still further protect the propeller, and a raking stern-post.

Second, peculiarities of construction; the unusual fastening, as noted above; the unusual and massive arrangement of the beams, and bracing of the sides to resist pressure; the introduction of screw tie rods to bind the ship together; the development of the ’tween-deck beams and waterways on a water-line, instead of on a sheer, like the upper-deck beams; the placing of the ceiling continuous from sister keelson to upper-deck clamps, and the placing of the ’tween-deck waterways, deck clamps, and the bilge strakes on top of the ceiling; the filling in of the bow almost solid where it meets the impact of the ice; the massive and unusual reinforcement of the rudder post to prevent twisting; the adoption of a lifting rudder, which may be raised out of danger from contact with the ice; the armouring of the stem and bows with heavy plates of steel; the protection of the outer planking with a 2–inch course of greenheart ice sheathing.

Peculiarities of rig are: pole masts throughout; very short bowsprit, which can be run inboard when navigating in ice of considerable elevation; three-masted schooner rig with large balloon staysails. The Roosevelt carries fourteen sails, including storm staysails, and has a sail area somewhat less than that of a three-masted coasting schooner of the same size.

Peculiarities of the machinery installation are: a compound engine of massive construction; an unusually heavy shaft of forged steel 12 inches in diameter; a massive propeller 11 feet in diameter, but with blades of large area, which are detachable in case of injury; a triple boiler battery; arrangements for admitting live steam to the low-pressure cylinder, in order to largely increase the power for a limited time; an elliptical cruiser-type smoke-stack to reduce wind resistance.

The best quality of material and labour were put into the ship, and it was believed and has since been proven that she is the ablest ship ever built for Arctic exploration.