From Cape Sabine the most practical course lies along the west shore, where at ebb-tide a navigable lane of water is often to be found between the shore ice and the moving pack. In 1905, after leaving Cape Sabine and working northward along the west shore past Bache Peninsula and Hayes Point, we were forced to seek shelter in Maury Bay to avoid the heavy ice advancing swiftly before a stiff northerly wind. By keeping a close watch on the ice and availing ourselves of every opportunity to advance, we followed the shore-line up past Scoresby Bay and Richardson Bay. Two attempts to reach Cape Joseph Goode failed, each time the Roosevelt being driven back to Cape Wilkes by the ice-pack. Rawlings Bay was packed with ice, and conditions to the northward, on the Grinnell Land side, altogether so unfavorable, that I determined to cross Kennedy Channel and proceed northward on the Greenland side, previous experience in this region having led me to believe that in most seasons Kennedy and Robeson channels could be more easily traversed on the Greenland side than on the Grinnell Land side.
After a long, hard struggle we reached the loose ice off Cape Calhoun, and headed north from Crozier and Franklin islands. Finding the channel which lies between Franklin Island and Cape Constitution impracticable, we followed the main channel close to Franklin Island.
As far as Joe Island it was fairly easy sailing as polar navigation goes. Making the Roosevelt fast to the ice-foot here, a trip to the summit of the island showed the Greenland side of Hall Basin as far as Cape Lupton, and possibly up to Cape Sumner, free from ice, while the Grinnell Land coast was filled with heavy ice, making navigation out of the question. Just beyond Cape Lupton, while breaking a way through a small gap in the ice, a quick change in the current, which runs very swiftly in this deep and narrow channel, forced the ice-floes together about the Roosevelt, smashing her up against and along the ice-foot. In less time than it takes to describe, it twisted the back of her rudder, snapped her tiller-rods, almost put her steering-gear out of commission permanently, and necessitated a stop of several days at Newman Bay to make repairs.
We had hoped that a lead across Robeson Channel to the neighborhood of Cape Union would make the return to the west side of the channel comparatively easy, but in this we were disappointed.
In 1908 the route of the Roosevelt from Cape Etah to Sabine and up the west coast of Kane Basin, past Victoria Head, was virtually the same as in 1905. This year, however, we found Kennedy Channel almost free from ice, and with no fog to delay, the Roosevelt steamed her way up the center of it, and broke all previous records by navigating the channel’s one hundred miles of length in one day.
Before reaching Robeson Channel we encountered ice and fog, and were once driven over to the east coast at Thank God Harbor in an attempt to find an opening in the pack. With this exception the Grinnell Land and Grant Land coasts of the channels were found practicable from Cape Sabine to Cape Sheridan.
On the return voyage from Cape Sheridan to Etah in 1908 I determined to try out a new route in these narrow and ice-filled channels. Instead of hugging the shore, the Roosevelt, on reaching Cape Union, was deliberately driven out into the pack-ice in order to work her way down the center of Robeson and Kennedy Channels. For a ship not specially built for ice-work such a course would be almost certain to result in disaster, but for one of the Roosevelt type, and in the hands of experienced ice-navigators, I consider this by far the preferable return route. It is also the quickest route, the trip from Cape Sheridan to Cape Sabine taking only twenty-three days, or twenty-three days less time than by the old route in 1906.
The navigation of polar waters demands incessant watchfulness and instant readiness even under apparently the most favorable conditions. During the passage of Kennedy and Robeson Channels Bartlett was nearly always in the crow’s-nest, and while I had almost unbounded faith in his judgment, I spent much of the time in the rigging below the crow’s-nest, watching the ice ahead, and in the worst places often relieving Bartlett of too great a load of responsibility by backing up his judgment with my own views. The periods of night at such times might as well not have been, for it is possible to get only snatches of sleep in the short times when nothing else remains to be done, and Bartlett and I have spent days and even weeks at a time in these regions without thinking of taking our clothes off to sleep.
The chief engineer, like his assistants, stood his eight- or twelve-hour watch, and was almost always to be found in the engine-room when the Roosevelt was passing through dangerous places; for any slip in the machinery at a critical time would have resulted in the loss of the ship.
The Roosevelt has undoubtedly deliberately struck heavier blows while fighting ice than any other ship would dare to attempt. Many times she has reared and risen on a steel blue mass of old floe-ice till I was reminded of a hunter rising to a stone wall. Repeated blows of her steel stem in the same spot have at times split pieces of floes, or the projecting tongue of a big floe which barred our passage, of almost incredible thickness just as a small hand ice-pick, if properly used, will split a large cake of ice.