Captain Lee P. Gillette, First Nebraska Cavalry, and the second under command, of Captain James B. David, Seventh Iowa Cavalry, and at once proceeded in a northerly direction, intending to strike Cedar or Calamus River that night. In this, however, I was disappointed, finding the country for twelve miles north one interminable mass of intricate hills and deep, waterworn gullies or canons, many of them being 200 feet deep, and compelling me to vary my course from due north to every point of the compass to effect a passage over them.

On arriving at the northern edge of those canons, which I named Elk Canons, from the great numbers of that animal seen in them, I was deterred from proceeding farther by seeing before me an interminable stretch of high sand-hill peaks to the north, separated from the Elk Canons by a small valley about three miles in width. Carefully examining the distant sand hills with a powerful field glass, and being satisfied from the lateness of the hour that I could not possibly reach water that day if I continued on, I countermarched the column back about eight miles to a slough we had passed, arriving there at 7 p. m., when we camped for the night. Early next morning we started forward toward Cedar River, reaching the southern line of the sand hills seen the evening before about 9 a. m., and traversing through them with great difficulty we came out upon another valley to the north of them at 12 m., reaching Cedar River at 1 p. m., which we found to be a mere creek at this point and utterly destitute of timber, save now and then a little isolated clump of willow switches. Turning up the south bank of the stream and marching westward we were repeatedly compelled to enter the sand hills to pass around elbows of Cedar River running bluff against the precipitous hills. At 3 p. m. we struck an Indian trail running south, about one month old, which is indicated on the map. Following this trail until 4. 30 p. m. we came to a point where the Indians had camped on the north bank of the river, and ascertained their number to have been about sixty or less. Here we camped, our stock being much exhausted in marching through soft sand in the hills during the forepart of the day. Early next morning we moved up the north bank of the river on the train, intending to follow it to the village from which it started. The road along this side of Cedar River was very good and level. We reached the head of the stream at 12 m., having marched twenty-eight miles that morning. There we found the trail coming in from the north, and smoke was thought to be seen to the northwest, but so indistinctly as to leave the matter doubtful. To the north and west the sand hills appeared to be less frequent, apparently jutting out of the level plain as isolated peaks, presenting no formidable obstacle to progress. Thinking it possible to reach the Niobrara River that night, I pushed to the northward at once, and reached the headwaters of one of the tributaries of the Niobrara about 7 p. m., having had a very fair road. While arranging camp smoke was observed a little north of west and distant about eighteen or twenty miles, as nearly as could be judged. Almost immediately after the appearance of this smoke other columns to the number of three were observed beyond it and to the northward. We all felt sanguine that we were now in the vicinity of hostile Indians. Sending a party out to examine the direction of the trail we had followed they returned about dark reporting that no further trace of the trail could be found, the ground being so hard as to prevent any impression. A terrific hail and thunder storm raged during part of this night, rendering the presence of every man with his horse necessary to prevent a stampede. At 3. 30 a. m. the following morning I pushed straight for the point where the