the pilot, by using the stern paddle wheel to cut through the silt, often backed the steamer upstream when it became impossible to proceed bow first. In case the barge became stranded, the steamer could be eased down close to her, and, by spinning the paddle wheel, a sufficiently strong current might be raised to wash her free.
Fuel for the "Arizona Fleet" was collected as the boat progressed upstream. Native mesquite and cottonwood timber growing along the river were the chief sources of supply, and fuel stations were established at intervals of twenty-five miles. Here the Indians were taught to cut and cord the wood into the required lengths, and for their services they were paid two dollars and fifty cents a cord at the fueling station.
The great seasonal variation in the volume of water on the Colorado created still another problem. During extremely dry periods as little as one thousand second feet have been measured,^^* whereas a peak flood measurement of 150,000 second feet was recorded in June 1909.^^^ In both extremes little could be done except wait for conditions to revert to normal. Because of the definite dry season, it was generally conceded that to operate successfully on the Colorado, a boat must of necessity be one of light draft, usually sixteen inches to two feet. To the desert pilot who seriously considered the possibilities of navigating an Arizona mirage, "Five feet was practically no bottom," and when the "leadsman called 'scant four' . . . the pilot did not snatch the boat back, reverse engines, pull bells, and shout down the tube to the engineer. He merely wiped his brow and reflected that the river must be rising."15«
The immediate effect of these many problems was to be seen in the impossibility of maintaining a definite sailing schedule on the Colorado. Schedules on the lower river between La Paz, Yuma, Ehrenberg, and Olive City were fairly dependable, but above these points, if arrival was made within a week or two of the time specified, conditions were considered average. The period of time occupied on a voyage depended on the good fortune of the pilot, the steamer, and the stage of the river. When the vessel was loaded, a speed of five or six miles an hour upstream was a good average. In 1872 the Sentinel reported that Captain Mellon with the steamer Cocopah made the round trip from Yuma to Mohave, a distance of 232 miles, in five days.^" In August, a year later, the Mohave covered the same run with a barge in tow in seventytwo hours, the "fastest time ever made on the Colorado River."^^^ Another exceptional run was made in September 1 873 from Yuma to the mouth of the river and return, a distance of 340 miles, in six days including stops.^^^
The first indication of a decline in steam navigation on the Colorado River came in 1876. Two years earlier, in December 1874, business was temporarily thrown off schedule when the steamer Montana was wrecked in the Gulf of California. She was towed back to San Francisco for repairs and was operating again by March 1875. Misfortune continued to stalk her. On De
