At midday on the 3d of February we left the Alatna River and took our way across country for the Kobuk. We had now no trail at all save what had been made a couple of months before by the only other party that had crossed the portage this winter, and it was buried under fifteen or sixteen inches of snow. There was quite a grade to be climbed to reach the plateau over which our course lay, and the men, with rope over the shoulder, had to help the dogs hauling at the sled. Indeed, over a good deal of this portage, from time to time, the men had to do dog work, for the country is rolling, one ridge succeeding another, and the loose, deep snow made heavy and slow going. One man must go ahead breaking trail, and that was generally my task, though when the route grew doubtful and the indications too faint for white man's eye, Roxy took my place and I took his gee pole, and slipped his rope around my chest. Breaking trail would not be so laborious if one could wear the large snow-shoes that are used for hunting. But the hunting shoe, though it carries the man without fatigue, does not help the dogs. The small shoe known as the trail shoe, packs the snow beneath it, and by the time the trail breaker has gone forward, then back again, and then forward once more, the snow is usually packed hard enough to give the dogs some footing. Footing the dog must have or he cannot pull; a dog wallowing in snow to his belly cannot exert much traction on the vehicle behind him. The notion of snow-shoeing as a sport always seems strange to us on the trail, for to us it is a laborious necessity and no sport at all. The trail breaker thus goes over most of the ground thrice, and when he is anxious at the same time to get a fairly accurate estimate by the pedometer of the distance travelled, he must constantly remember to upend the instrument in his pocket when he retraces his steps, and restore it to its recording position when he attacks unbroken snow again. Also he must take himself unawares, so to speak, from time to time, and check the length of his stride with the tape measure and alter the step index as the varying surfaces passed over require. Conscientiously used, with due regard to its limitations, the pedometer will give a fair approximation of the length of a journey, but a man can no more tell how far he has gone by merely hanging a pedometer in his pocket than he can tell the height above sea-level of an inland mountain by merely carrying an aneroid barometer to the top. [Sidenote: THE SUNRISE AND THE MOUNTAINS]