[I-57] The total of 12 m. does not agree with the text, which calls for 11 + 4½ = 15½. Eagle cr. seems to have been struck about on the boundary between Coffey and Lyon cos., where Four Mile cr. falls into it. It is a considerable stream, which heads in the divide about Olpe (a place on the A. T. and S. Fé R. R.), is increased by Harper, Hoosier, and other tributaries, and runs E. into the Neosho a mile and a half above Strawn (a place on the Mo. Pac. R. R.). To reach Eagle cr. from North Big cr. Pike passed opposite Burlington, seat of Coffey Co., several miles to his right, and headed the small Otter cr., on or near which is a place called Patmos.

[I-58] It may not be possible to decide which of the several branches of Eagle cr. Pike went up to the divide. To send him up the main branch, past Olpe, agrees best with his 21 m. to-day; but in that case he must have breakfasted late. There is a sharp elbow in his dotted trail, which would seem to indicate that he made a turn from his former course over the divide. Aside from any questions of detail, which perhaps could not be decided even by a resident of the region traversed, we have Pike safe on the headwaters of Vermilion or Verdigris r. (it has these alternative names on recent maps). It heads in the divide which Pike has crossed, by numerous small tributaries, several of which Pike charts. Among them are Haldemand and Tate, heading opposite branches of Eagle cr., and further on Moon, Rock, Fawn, and Camp creeks. The Verdigris is of a size smaller than the Neosho, W. of which it runs in an approximately parallel course; it leaves Kansas through Montgomery Co., enters Indian Territory, and in the country of the Creek Indians falls into the Arkansaw 8 m. above the mouth of the Neosho. Pike lays it down well, especially the fan-shaped leash of branches in which it heads, but runs it into the Arkansaw in common with the Neosho. The Verdigris has of course its proper basin or drainage within the more general watershed of the Neosho and other Arkansan as distinguished from Missourian waters. The rim of this basin is the divide Pike crosses over to-day. He camps on one of the small headwaters, probably Fawn or Camp cr., in the close vicinity of the places called Elco and Verdigris.

[I-59] Pike has headed Verdigris r., and recrossed the brim of its basin into the Neosho basin again. In cutting off this small segment of the Verdigris basin he passed from Lyon into Chase Co., "over high hilly prairies," i. e., the divide, and continued westward till he struck "a large branch of Grand r." We discover later that Pike takes Cottonwood r. to be the main Grand, i. e., Neosho r., which I do not see that it is not, though the other one retains the name of Neosho above their confluence. The stream he strikes is the S. fork of Cottonwood r. This heads in the same hilly country, by tributaries known as Little Cedar, Thurman, and Mercer creeks, in relation or opposition to the uttermost sources of Verdigris r., and flows N. to fall into the main Cottonwood 4 m. below Cottonwood Falls, county seat of Chase. Pike probably came on this stream somewhere in the vicinity of Baker or Crocker cr., between places called Matfield Green and Bazar.

Cottonwood "creek" was originally so named at the point where it was struck on the old Santa Fé caravan road, and because it showed the first trees of that kind to be found in traveling westward on that route. The crossing was at or near present town of Durham, Marion Co. It was some time before the true connection of the Cottonwood with the Neosho was made out. At Council Grove the traders knew they were on a head of the Neosho or Grand r., though they called it Council Grove cr. They kept on W. to "Diamond springs" (on a head of Six Mile or present Diamond cr.), and thence to "Lost spring" on their "Willow" cr. (a head of present Clear cr., which falls into the Cottonwood at Marion); the next stream they struck being this Cottonwood cr., at or near Durham: see a note beyond, where I undertake, perhaps rashly, to recover the old caravan road in terms of modern geography.

[I-60] If Pike bore as much N. of W. as his dotted trail seems to indicate, the mileage would fetch him on Cottonwood r. about the situation of Cedar Grove and Cedar Point, which are within a mile or two of each other and of the boundary between Chase and Marion cos., and about 6 m. down river from Florence, Marion Co. He is evidently in the loop which the Cottonwood makes S. E. from Marion to Florence, and then gradually N. E. to the vicinity of Cottonwood Falls. If the old Kansas Indian trail the map lays down could be recovered or identified, it would serve to locate him still more precisely. He crosses the Cottonwood and camps on its left bank. If we attentively regard the camp-marks of the 12th and 13th, we find them close together, N. of the Cottonwood, S. of a creek flowing E., and W. of a pair of creeks flowing S. These requirements are fulfilled, if we take the one running E. to be Middle creek, which falls in by Elmdale, 10 m. below Cedar Grove; and the other two, those that fall in together at Marion, 12 m. (direct) above Cedar Grove. It is true there are several creeks nearer, on the same side, as Silver, Bruno, and Martin, but these are all smaller than such as Pike usually charts, and, moreover, he could not go his 9 m. to-morrow in any direction without getting beyond them.

[I-61] Camp in the close vicinity of Marion, seat of the county of that name. The Indian trail seems to have run past or through Marion. We can confidently locate Pike within 3 m. of the town on the night of the 13th; and Marion thus furnishes an excellent fixed point whence to trace him on to Smoky Hill r. The two streams which unite at Marion, and run through the place as one, are called Brook Luta and Clear cr.

[I-62] Continuing past Marion, up the Cottonwood, which he has to his left, Pike camps near Durham, Marion Co. This town is on the river, and through it runs the Chic., Kas. and Neb. R. R. The route seems to have sheered off from the river a little to the right, more in line with Brook Luta than with the Cottonwood itself: see next note.

[I-63] Passing north between Cottonwood r. on his left and Brook Luta on his right, Pike makes the divide in the vicinity of Tampa, Marion Co. This is a village on the head of Brook Luta; the railroad last named goes through it, and Pike crosses the line of this railroad between Durham and Tampa. He is flanking the higher hills (1,500 feet or more) in which the main Cottonwood heads, by leaving them to the left or W. This is a somewhat roundabout way from the vicinity of Durham to that of Bridgeport on the Smoky Hill, where Pike strikes this river early on the 17th; but it is evident that he did not go straight between these points, for they are only about 25 m. (direct) apart, and we have to account for 18 + 13 + 6 = 37 m. on the 15th, 16th, and morning of the 17th. These mileages adjust themselves to a nicety by the way I make out. I suppose he crossed the divide between Tampa and Kuhnbrook, Marion Co., thus passing from Arkansan to Missourian waters, as he says. Kuhnbrook is a little place on one of the heads of the large Turkey cr., which runs N. into the Smoky Hill r. opp. Abilene. Rhoades is the next place on this branch of Turkey cr., and in passing to its vicinity Pike crosses from Marion into Dickinson Co. He continues on, bearing to the vicinity of Elmo and Banner. These are places near another head of Turkey cr., and both on the Mo. Pac. R. R.; they are within a mile of each other. Pike keeps on a piece, westerly, and sets camp in the vicinity of Carlton, Dickinson Co. Carlton is between the two forks of Holland cr., next W. of Turkey cr., with which Holland runs parallel to fall into the Smoky Hill r. opp. Abilene. Carleton is 7 m. due E. of Gypsum City, which latter is on a creek of that name Pike next strikes.

[I-64] Pike camps on a branch of Gypsum creek. This is a large stream which heads in close relation with the uttermost sources of Cottonwood r., in the vicinity of Canton, McPherson Co., and flows due N. into the Smoky Hill, between the mouths of Solomon and Saline rivers. It is larger than either Turkey or Holland cr., and much branched. It runs about halfway between Holland cr. and the Smoky Hill, parallel with both; for the latter, having made a bold sweep from the W., curves N. past Lindsburg and Bridgeport to Salina, and thence E. to receive first the Saline, next Gypsum, and then Solomon r. On Gypsum cr. are Chico and Gypsum City, 10 and 12 m. above its mouth; and Pike strikes it a few miles further up or S. of these towns. Pike charts its headwaters elaborately, and sets his camps of the 15th and 16th among the five branches he lays down. Probably one of these should be taken for Holland cr.; the other four are less easily identified. From his position in the vicinity of Carlton Pike passes W. from Dickinson into Saline Co., comes first to Hobbs cr., next upon Gypsum itself and Stag cr., in quick succession; crossing these three he continues W. to another branch of Gypsum cr., namely, that one now meandered by the Mo. Pac. R. R. between Gypsum City and Bridgeport. He camps on the latter, 6 m. E. of the Smoky Hill.

[I-65] "We passed it six miles to a small branch to breakfast" is a dubious phrase which I understand to mean that Pike went 6 m. to a small branch to breakfast, and then crossed the Smoky Hill r. at once—at nine o'clock. This crossing was in the immediate vicinity of Bridgeport, and perhaps at the very place the Council Grove, Smoky Valley and Western branch of the Mo. Pac. R. R. crosses to run into Bridgeport. Two insignificant runs fall into the river from the east within 3 m. below Bridgeport; the first of them is named Pawnee cr. Crossing the river, Pike proceeds up it, but a little W. of N., and bearing somewhat away from it; he passes Dry cr., which lower down runs through the county seat Salina, and camps on Mulberry cr., 2 or 3 m. due W. of that city, and about the same distance below the point where Spring cr. falls into Mulberry. This stream skirts north of the city, receives Dry cr., and falls into Saline r. a mile or two further. Salina is a large place, one of the best known in the State, where four great railroad lines meet—the U. P., Mo. P., A., T. and S. F., and C., R. I. and P. Six or 8 m. due W. of the place where Pike crossed the river are the Smoky hills, or Smoky Hill Buttes, celebrated in story if not in song. The great river named from these conspicuous landmarks is the main southern fork of Kansas r., as the Republican is the northern. Its uttermost sources are in Colorado. Receiving uncounted tributaries in its long course, it runs E. in Kansas through Wallace, Logan, Gove, Trego, Ellis, Russell, and Ellsworth, loops S. into McPherson and out again N. into Saline Co., makes an elbow at Salina and continues E. through Dickinson into Geary Co., where it joins its mate between Junction City and Fort Riley, thus composing the Kansas. Two of its largest branches are the Saline and Solomon.

This finishes the first section of Pike's Kansas route from the Little Osage to Salina. The rest of the way to the Pawnee Republic is northward, crossing in succession Saline r., Salt cr., Solomon r., Buffalo cr., and White Rock cr., striking the Republican r. in Webster Co., Neb., near the S. border of that State. The distance is less than the 97 m. Pike makes of it. His map is extremely faulty; he seems to have gone about N. W., though his actual route was very little W. of N. It also runs Saline and Solomon rivers far apart into the Republican, instead of the Smoky Hill, magnifies Salt cr. out of all proportion, and minimizes both Buffalo and White Rock cr.

As a bit of authentic history which may interest those in Salina who have reason to be proud of the growth of their city during one generation, I will transcribe a passage from my own field notebook, made when I was staging from Leavenworth to Santa Fe, in 1864: "Sunday, May 29th. Left Junction City and came to a place called Salina—three houses and a pig stye."

Fort Riley, as above mentioned, was begun by Major Edmund Augustus Ogden, who had selected the site and was occupied with the work when he died there Aug. 3d, 1855, in the epidemic of cholera then raging. He was born at Catskill, N. Y., Feb. 20th, 1810; removed to Unadilla, N. Y., and from there entered West Point July 1st, 1827; he became second lieutenant of the 1st Inf., July 1st, 1831; first lieutenant, Dec. 17th, 1836; was transferred to the 8th Inf., July 7th, 1838; promoted to be captain, Dec. 1st, 1839; and was breveted major for meritorious conduct. His first duty was at Prairie du Chien; his marriage with Captain Gustavus Loomis' daughter Eliza, at Fort Snelling, is said to have been the first ceremony of the kind between white persons in Minnesota; he served faithfully and with distinction in the Black Hawk, Florida, and Mexican wars, and for many years discharged arduous and responsible duties in the quartermaster department. For several years immediately preceding his death he was stationed at Fort Leavenworth.

[I-66] Saline r., distinctively called Great or Grand Saline, has been already noted. Pike crosses this, and proceeds to "a small dry branch" of the next river, to camp for three days. This river is the one he calls Little Saline, and is now known as Covert or Salt cr., a branch of Solomon r. which falls into the Solomon 4 m. below Minneapolis, Ottawa Co. Pike's map connects it correctly, but magnifies its size; for the stream which he passed on returning from the Pawnee village, and which he lays down as a head of his Little Saline, is a branch of the Great Saline. Pike probably crossed Saline r. in the vicinity of Culver, where the railroad now does, then soon passing from Saline into Ottawa. The small branch of Salt cr. on which he camped was one of several such in the vicinity of Ada.

[I-67] An error is here evident, and I suspect some confusion of the diary of the 21st and 22d. 1. "Distance 11 miles," for the whole 22d, is necessarily wrong, if m. were made in the afternoon, and this, too, after marching from eight to eleven in the forenoon. 2. Aside from the fact that there is no branch of the Republican fork in this vicinity, the map shows that Pike did not reach Solomon r. till the 23d, and the text of that day confirms this. Camps of the 21st and 22d were both in the space traversed between Salt cr. and Solomon r., less than 20 m. at the furthest. 3. The difficulty disappears if for "12 miles," etc., of the above questionable clause, we read "2 miles to the first branch of Solomon river on our route." This would set Pike on one of the small creeks that fall into the right bank of Solomon r. in the vicinity of Glasco, Simpson, and Asherville. A former name of Solomon r. was Nepeholla, used, e. g., by Gunnison, P. R. R. Rep. II. 1855, p. 17. Capt. J. W. Gunnison came to the mouth of Solomon's fork July 6th, 1853, from Westport, Mo., by the Wakarusa River route, striking the Kansas r. at Fort Riley, crossing there, and continuing through Abilene; he was en route to the great bend of the Arkansaw by the usual Smoky Hill route.

[I-68] Taking Pike northward across Buffalo cr. I suppose this was crossed somewhere in the vicinity of Jamestown, Republic Co. In this position the Republican r. itself is only 5 or 6 m. to his right, and the rest of the journey is simply following up this river obliquely on about a N. W. course, at a somewhat increasing distance from it, until he nears it to approach the village.

[I-69] To White Rock cr., west of White Rock, a town on the creek and on the boundary between Republic and Jewell cos. This stream runs east through these, and falls into the Republican r. opposite Republic City. In getting here, Pike seems by his map to have crossed several small streams running to his left, and into a stream he runs into Solomon r. I suppose these to be some branches of Marsh cr., a sluggish tributary of Buffalo cr. from the N. W.

[I-70] Finishing the journey to the Pawnee Republic village, whence the great river on which it was situated took its name. Its ultimate sources are in Colorado, like those of the Smoky Hill r. Its main course then cuts off the extreme N. W. corner of Kansas, by running through Cheyenne Co.; whereupon the stream enters Nebraska, and skirts the southern border of this until it dips into Kansas across the N. border of Jewell Co., whence it continues E. into Republic Co., turns S. in this to Cloud Co., E. through this to Clay Co., and S. E. through this to Geary Co., where it is joined by the Smoky Hill, as already noted. The whole journey thus made from the Osages to the Pawnees foots up, by Pike's distances, about 350 m. In a letter to the Secretary of War he calls it "375"; but this is simply offhand. He also claims that his Osages led him roundabout 100 m. through their fear of the "Kans." Pike's land mileages seem to me more correct than those excessive ones he assigns to his navigation. I suppose this journey to have been between 300 and 325 m.

I must emphasize here the fact that I have failed in every attempt to locate the precise site of the Pawnee village. One would suppose it well known; I find that it is not, and have yet to discover the ethnographer or geographer who can point it out. Correspondence addressed to persons now living in the vicinity was as fruitless as my exploration of the sources of official knowledge in Washington, where several friends interested themselves in my behalf to no purpose. I know of no closer indication than that afforded by Gregg's map of 1844. This letters "Old Pawnee Village" on the S. bank of the Republican, halfway between long. 98° and 99° W., and thus, as I judge, about opposite the present town of Red Cloud, Webster Co., Neb. Gregg runs the Republican entirely S. of lat. 40° N., i. e., in Kansas; but the place where Pike struck it was certainly in that portion of its course which runs in Nebraska, just over the Kansas line. Gregg in fact gives his river a recognizable northward convexity along here, and if it does not overreach 40°, that is a fault of absolute, not relative, position. We are here much less concerned with latitude than with longitude. The river is running approximately from W. to E., in Webster Co., and the main point is how far W. the village was, as that would affect details of the route from the last point at which I have been able to locate Pike. It will be necessary to discover the exact situation of the Pawnee village before the cloud over the end of this journey can be dispelled, and the beginning of the journey from the village to Great Bend on the Arkansaw can be set in a clear light. For the present I can only tentatively assume longitude 98° 30´ W. (See Scandia, in the Index.)

[I-71] Guillaume Thomas François Raynal, commonly called Abbé Raynal, b. Aveyron, France, Apr. 12th, 1713, d. Paris, Mar. 6th, 1796—a philosophical free-thinker and historian, who wrote too much sense and truth to suit his official superiors, and was consequently unfrocked. It is a curious fact in the history of the most Tammany-like machine for the propagation of painful superstitions ever known in the Western world—excepting perhaps Brigham Young's similarly organized scheme—that whenever one of its members begins to think for himself they make him take off his gown and wear trousers openly. The irony in the case seems to escape the professional nurserymen in that hot-house. The abbé wrote various works; his most celebrated one, to which Pike refers, is: Histoire Philosophique et Politique des Établissements et du Commerce des Européens dans les Deux Indes, 1770, repub. 1780-85—a book whose strength and other merits may be inferred from the fact that it had the honor of being burned by Parliamentary order; though its author was simply exiled, the times being already a little out of joint for roasting heretics along with their heresies.

[I-72] Richard Sparks of Pennsylvania had been a captain in the levies of 1791, when he was appointed a captain of infantry, March 7th, 1792; he was arranged to the 3d Sub-legion, Sept. 4th, 1792, to the 3d Infantry, Nov. 1st, 1796, and transferred to the 2d Infantry Apr. 1st, 1802; he became major July 29th, 1806, lieutenant-colonel Dec. 9th, 1807, and colonel July 6th, 1812; he was honorably discharged June 15th, 1815, and died July 1st, 1815.

[I-73] 1. As already indicated, "Tetaus" and "Tetans" are Pike's names for Comanches, also variously known as Ietans, Jetans, Hietans, Aiatans, etc., and also by the Sioux name Padoucas, adopted by the French; they called themselves Num, meaning simply "people." Some of their other names are Kaumains, Choumans and Comandes; we now write Comanches or Camanches indifferently, thus adopting a form of the Spanish name, whose meaning is unknown. These Indians are of the Shoshonean family; they number about 150, on the Kiowa, Comanche, and Wichita Reservation in Oklahoma; there were some 2,500 when they were placed in a reservation in 1868; they had been noted, time out of mind, as wide-ranging, lawless, and warlike freebooters. 2. Pike above mentions three of the four principal tribes of the Pawnee confederation, i. e., of the middle group of Caddoan stock, who are: (1) Pawnee proper, Grand Pawnee, or Tcawi; (2) Pawnee Republicans or Republican Pawnees (giving name to the great branch of the Kansas r.); (3) Pawnee Loups, Pawnee Mahas, Pawnee Wolves, or Skidis; (4) Tapage or Pitahauerat: see further, L. and C., ed. 1893, pp. 55-57. (3) Pike's Kans are entirely different Indians, of Siouan stock, Dhegiha group: for these see l. c., pp. 33, 34.

[II-1] On the other hand, Lewis and Clark first heard of Pike's expedition on Sept. 10th, 1806, when they were nearing the Big Nemaha on their way down the Missouri, and met a boat with a trader bound for the Pawnee Loups: see L. and C., p. 1206.

[II-2] Letters to Generals Dearborn and Wilkinson, sent by this express, formed Docs. Nos. 13 and 14 in the App. to Pt. 2. of the orig. ed. They are given beyond.

[II-3] Camp is on one of the tributaries of Rock cr., close to that of Sept. 24th, if not on the same spot.

The route now taken by the expedition is very little W. of S., to strike the Arkansaw r. at the most convenient point. Thus it diverges westerly from the route by which the Pawnee Republic was approached, which was W. of N. The two make a ᴧ whose legs rest on the Smoky Hill fork at the two points where this was crossed in going and returning, with the apex at the village. The main streams crossed between the Republican and Smoky Hill forks are Solomon and Great Saline rivers. Pike is also on the trail of the Spaniards who have just raided United States territory to the Pawnees; he marks their camps, as far as he can find them out, with a o, to distinguish them from his own, marked ×.

The party which leaves the Pawnees, so far as the white men are concerned, only differs from that which left Belle Fontaine by the absence of the deserter, Kennerman: see note², p. 358. The express which Pike dispatched therefore consisted of some of his Indians.

[II-4] This close calculation was doubtless based in part on information Pike already possessed. We have been told that Malgares started on his raid with 100 dragoons and 500 militia, of which 600 men 240 had been detached, leaving 360. The "large branch" on which was camp was probably one of the heads of Livingston cr.

[II-5] "Distance 18 miles" would never bring Pike to any branch of the Solomon. His error here is a puzzling one until it is detected by reference to the map. That sets his camp-mark high up on the same creek, several branches of which he had on his left when he went up E. of it, Sept. 24th. It is Buffalo cr., which Pike erroneously runs into Solomon r., and so seems never to have passed before. See back, note⁶⁸. I suppose he struck Buffalo cr. below Mankato, somewhere in the vicinity of Jewell and very likely between these places. He seems to be holding about S. S. W., and to-morrow strikes Solomon r. But, as already explained, note⁷⁰, p. 410, precision in this matter is impossible, without knowing exactly where the Pawnee village stood, so as to have a fixed initial point of the journey. I understand that there was a certain "Pawnee trail" once well known from this village to Great Bend on the Arkansaw. If this be now determinable, it will represent Pike's route with a closer approximation to accuracy than I have been able to follow it out.

[II-6] To the Solomon r., in the vicinity of Beloit, Mitchell Co. The stream which Pike lays down across his trail of to-day is perhaps Plum cr., which falls in below Beloit.

[II-7] Hiatus in the text, probably from missing or illegible MS.; no course or distance for to-day. But the map shows a march, and sets camp among the heads of a small stream. This is perhaps Salt cr., high up, somewhere in the vicinity of Saltville, Paris, or Victor.

[II-8] Perhaps to vicinity of Lincoln, seat of the county so called. The map has an extra camp-mark, on the head of what Pike calls "Little Saline river."

[II-9] In saying that he crossed the (Grand Saline) river "two or three times," Pike does not mean that he meandered that stream on his march, but that he or some of his party were hunting about for the Spanish trail which he was so eager to follow, and which here became blind. His map marks o o, the two Spanish camps he found. His was on the north bank of Smoky Hill r., whose other name in the text, "La Touche de la Cote Bucanieus," possibly stands for La Fourche de la Côte du Kansas, i. e., that fork of the Kansas which runs along the dividing ridge or coteau—which is perfectly true of the Smoky Hill fork. Pike struck the Smoky Hill in Russell or Ellsworth Co. Camp of the 13th is about on the border of Russell and Barton cos., in the vicinity of Forest Hill and Dubuque.

[II-10] The approximation of Missourian and Arkansan waters is here very close. The Arkansaw makes its great bend northward into Barton Co., whose county seat is named Great Bend accordingly. The courses of the Smoky Hill and Arkansaw are for many miles approximately parallel, and only some 30 m. apart in air-line distance; the numerous tributaries of each arise all along the ridge which forms the divide between these waters. Pike has crossed the divide, and is now on one of the headwaters of Cow cr., a large affluent of the Arkansaw, which traverses Barton and Rice cos. in a southeasterly course, and falls in at Hutchinson, Reno Co. His camp appears to have been somewhere in the vicinity of Claflin, Barton Co., on the Kas. and Col. R. R. The stream is laid down on his map. It is by far the largest tributary of the Arkansaw between the Little Arkansaw and Walnut creeks. It was the last stream to be crossed on the old Santa Fé caravan road before the Arkansaw was reached. This road also crossed the several tributaries of Cow cr. in the vicinity of Lyons, Rice Co. One of these, between Lyons and Chase, was called Little Cow cr. We find another, E. of Lyons, marked on modern maps as "Jarvis" cr., and given as Charez or Owl cr. in Beckwith's Report of 1853, P. R. R. Rep. II. 1855, p. 22. Two of these names refer to Don Antonio José Chavez, who left Santa Fé in February, 1843, en route for Independence, Mo., but was brutally murdered and robbed in this vicinity, on or about April 12th, by a party of 15 men who represented themselves to be Texan troops under the command of one John McDaniel. Particulars of this outrage are given by Gregg, Comm. Pra. II. 1844, pp. 166-169.

[II-11] Walnut (Big or Wet Walnut) cr. is that large northern affluent of the Arkansaw which runs E. from Lane through Ness and Rush into Barton Co., and falls into the river 4 m. below Great Bend, county seat of Barton. A branch of this, called Little or Dry Walnut cr., runs E. from Rush into Barton, and falls into Walnut cr. about 4 m. from the mouth of the latter. Great Bend is on the N. bank of the Arkansaw, and thus between that river and Little Walnut cr. The way in which, and the precise point at which, the Expedition struck the Arkansaw could hardly be discovered from the text of the 15th-18th; we are not even told till the 18th that we are on the Arkansaw, as the 15th mostly, and the 16th and 17th entirely, are taken up with the wanderings of the lieutenant and doctor, who got lost. The key to the situation is not found till the 23d, when it is luckily recited that a trip was made from the camp on the Arkansaw "about 20 miles to a large branch [or fork] on the right." This is the well-known Pawnee fork of the Arkansaw, where was old Fort Larned, a noted place, and where is now Larned, seat of Pawnee Co. So the Expedition struck the Arkansaw 20 m. below Larned, in the very suburbs of the present city of Great Bend. This locality about the mouth of Walnut cr. became early noted, not only as the place of northernmost deflection of the Arkansaw, but also as the first objective point on that river, where the old Santa Fé caravan road struck that river. It also became the site of Fort Zara, or Zarah—to be found on some maps as Fort Sarah—which was built in 1853 on the high ground between Walnut and Cow creeks, about 5 m. N. of the road. On July 12th of that year, Capt. Gunnison reached the great bend by the Smoky Hill route from Fort Riley, having been preceded in arriving there three days by his companion, who came over the regular Santa Fé route; Lieut.-Col. E. V. Sumner, 1st Dragoons, and other officers, arrived from Mexico the same night; and on the spot was camped Captain and Bvt.-Major Edward Johnson, 6th Infantry, about to build the fort, as that 100 m. further up the Arkansaw (Fort Atkinson) was to be abandoned. Col. John Garland of the 8th Infantry passed by in July of that year. Pike's approach was: Being in camp of the 14th on some head of Cow cr., the Expedition started on the 15th at noon, and marched five hours, about 15 m., on a W. S. W. course, thus crossing the Cheyenne Bottoms above said, and coming to Walnut cr. just above the mouth of the Little Walnut above described. Pike pointed out a wood and told Wilkinson to go there to camp, while he and the doctor would go up Walnut cr. a piece to hunt for the Spanish trail. Either mistaking the wood intended, or finding himself so near the Arkansaw, Wilkinson went on to that river and camped the party on its north bank, a mile or two above where Great Bend now stands. Pike and the doctor went shooting buffalo, and it got pretty late; they returned to where Pike had told Wilkinson to camp, and found nobody there; so they bivouacked on the spot. In the morning they went up Little Walnut cr. to search, but did not go far from those two buffalo they had killed; in fact they got rattled at finding no camp, turned about and went down Little Walnut cr. to its mouth (which is what text of the 16th means by "their junction"—confluence of the two creeks). On the morning of the 17th, being thoroughly alarmed, and imagining that the party must be higher up the Little Walnut, they started up again, but probably went a very little way in the rain; for they were overtaken early on the 18th by two men whom Wilkinson had sent in search of them, and then they were only "about three" miles from the camp on the Arkansaw. It is not likely they were at any moment 10 m. from the spot where they had left the party.

Pike's map shows nothing but the trail of the party, no camp being marked after that of the 12th, on the other side of the Smoky Hill r. The trail makes a sharp elbow at the point where, having come down Cow cr. on the 14th, they turned from that stream on the 15th. Besides Cow cr., three others appear in succession to the W. The first is Walnut cr.; the second is Little Walnut, a branch of the first, run separately into the Arkansaw; while the third is Ash cr., which falls in above camp. Cow cr. is brought in too near the next one. On the south side of the Arkansaw is marked the station of the 19th-27th, with the legend: "Here we struck the Arkansaw from whence L^t. Wilkinson descended the river in skin canoes and Capt. Pike went up by land with his party." This ends map I. of the Arkansaw, etc., and map II. of the same connects at this point, the first stream laid down being Pawnee fork, and the first camp that of the 29th. Camp of the 28th falls between the two maps, and is not shown. The Spanish trail, which Pike lost on Smoky Hill r., was all the while a little to the W. or right of the party, and is recovered on the S. side of the Arkansaw, on the 30th.

Pike elsewhere says of his journey from the Pawnees to the Arkansaw that it was on a general course S. 10° W. 150 m., but might have been made in 120. His deviation from the most direct route was in bearing a little too far W. to cross the Saline and Smoky Hill, and then some needless meandering across the divide to the Arkansaw. But he struck the latter exactly at the right point; for Great Bend is where the old Smoky Hill and Cimarron route from Leavenworth to Santa Fé reached the Arkansaw. There was of course nothing on the spot in Pike's time—nor was there even in 1864, when I first passed the place, excepting a miserable shack the stage company had built. The nearest settlement at that time was Fort Larned. My journal of May 31st, 1864, refreshes my memory: "At 2 p. m. we brought up at Fort Larned—mean place, built of adobe and logs, with a drunken officer in command; everybody half drunk already; and all were whole drunk by bed-time."

[II-12] Doted or unsound: see L. and C., ed. 1893, p. 951.

[II-13] Pawnee fork is larger than Walnut cr. It runs through several counties on a general E. course, and falls in at Larned, seat of Pawnee Co. When I was in the country, 30 years ago, the three principal branches were called Heth's, Buckner's, and Shaff's. A branch now rejoices in the name of Guzzler's Gulch. Saw-mill cr. is a long but slight tributary which falls in close to the mouth of the main stream. Pike crosses the mouth of Pawnee fork on the 29th; the Spaniards had crossed it higher up. He lays it down as a short, forked stream. Larned is now a city of some importance, and a rival of Great Bend; it is the natural development of which old Fort Larned was the germ; it is built mainly on the N. or left bank of Pawnee fork, but has lately crossed that stream, and also extended in the adjoining Arkansaw bottom. The locality became noted with the establishment of the Santa Fé trade in the '20's, and later on was a point of strategic importance in our relations with hostile or unruly Indians. The main road passed here en route for Santa Fé, in continuation both of the earliest caravan road and of the later Smoky Hill stage route; it offered a good camping place, which traders, troops, and other travelers generally occupied. Another reason for stopping was that the river was not easy to cross when full. Thus, when Emory and Abert were here, July 13th, 1846, one of Kearny's expressmen, A. E. Hughes, was drowned in it (J. T. Hughes, Doniphan's Exp., 1887, p. 21). But it varied much; July 13th, 1853, Gunnison and Beckwith found it 20 feet wide, with a fair current, and a depth of only a foot or two.

[II-14] This is an early but not the first account of the animals, and has been much cited, particularly as authority for the name wishtonwish (which J. Fenimore Cooper misapplied to the whippoorwill in one of his novels). The date of Pike's observation is subsequent to that of Lewis and Clark, but its publication was prior by four years; both these notices are antedated by Gass, 1807: see L. and C., ed. 1893, p. 111.

[II-15] A letter which Lieutenant Williamson bore to his father from Pike formed Doc. No. 15 of the App. to Pt. 2. of the orig. ed., and is given beyond in its proper place.

[II-16] The five soldiers who descended the Arkansaw with Lieutenant Wilkinson were: Sergeant Ballenger; Privates Boley, Bradley, Huddleston, Wilson. Lieutenant Wilkinson's separate report of his journey hence to the Arkansaw Post formed one of the Documents of the App. to Pt. 2 of the orig. ed., and will be found beyond, where it is annotated in due course.

Those who proceeded to the horrors of the mountains in midwinter and subsequent capture by the Spaniards were: Captain Pike; Dr. Robinson; Interpreter Vasquez; Sergeant Meek; Corporal Jackson; Privates Brown, Carter, Dougherty, Gorden, Menaugh, Miller, Mountjoy, Roy, Smith, Sparks, Stoute—16 all told: compare date of Oct. 7th, p. 419 and note², p. 360.

Pike now starts up the Arkansaw, to which he holds till he reaches the site of Pueblo, Col.

[II-17] Taking the party past Pawnee rock and the mouth of Ash cr., to a point about midway between the latter and the mouth of Pawnee fork. They traveled on the left or N. side of the river, approximately along the track of the A., T. and S. F. R. R., passing Dundee station and the small town of Pawnee Rock; Hubbard cr., on the other side of the river, is also passed, and camp is set a little beyond it, over the border of Barton Co., in Pawnee Co. The town of Pawnee Rock takes its name from the remarkable natural object of the same designation, also sometimes called Painted rock, which was a great landmark in old times. This is the most prominent point of a sandstone ridge of notably reddish color and in part scoriaceous; it is about 20 feet high, and stands off to the right of the road as you go up—about 2 m. from the Arkansaw r., before you come to the crossing of Ash cr. It was a convenient place for the Indians to exercise their pictographic art, and when the road came to be traveled by the whites the rock was soon covered with inscriptions of names, dates, and the like. It is about 9 m. by the road from the town of Pawnee Rock to the crossing of Pawnee fork.

[II-18] Passing Pawnee fork and Larned, Pawnee Co., to camp on the left or N. W. bank of the Arkansaw, about 5 m. beyond. Here is the place where the old Santa Fé road forked, in the days of the caravans and stages. The main road followed up the Arkansaw; but the right-hand road sheered off from the river to take up what was known as the "dry route"—a sort of cut-off which looked promising and became a regular stage-road, but was no great advantage when you had to go slowly and camp out, as the lesser distance was offset by lack of wood at all times, and of water at most seasons. Having been over this road, I can certify to the remarks of Gunnison and Beckwith, P. R. R. Rep. II., 1855, p. 24: "Five miles from camp [on Pawnee Fork] the road forks ... and one branch follows near by the windings of the Arkansas, to secure grass and water, while the other appears to push off for a 'short cut' and 'dry route' to Fort Atkinson, near which they again unite on the Arkansas river; but this appearance is deceptive; for after going a few miles it turns abruptly southward, and follows but a few miles from, and parallel with, the other road, keeping it generally in sight, as it does also the trees and sand-hills upon the banks of the Arkansas river, and is, except in the rainy season, without good grass and badly watered." The air-line distance of the "dry route," from the point where Pike is now to Dodge City, is about 54 m.; the actual travel is nearer 60. The ground passed over is that sometimes watered by the Coon creeks, and the road coincides to some extent with that now traversed by the A., T. and S. F. R. R. Of late the face of the country has been modified by the Eureka Irrigating Canal, which starts from the Arkansaw at Ingalls, hugs the river more or less closely to the bluffs below Dodge City, and then starts off across country in the direction of Spearville and Kinsley.

[II-19] Pike camps to-night about opposite Garfield, a railroad station and small village on the left or N. W. bank of the river. He started up on that side (having the river to his left), but crossed over on the 30th, and will continue the whole way to Pueblo up the right bank, having the river on his right. The general course of the river being from W. to E., its right bank is on the S., and thus N. of Pike.

[II-20] Kinsley, county seat of Edwards, is something of a town in these parts, situated a mile or two W. of the river on that one of the Coon creeks which runs oftener than the other one does, and which, when it has any water to discharge, falls into the Arkansaw at Garfield, after skirting the river for many miles. The nomenclature of Big and Little Coon creeks is reversed on some maps. I find that I was camped on one of them, 24 m. from Fort Larned, June 1st, 1864, under which date my old journal calls it "a puddlesome slough on the prairie." Thirty years ago it was good buffalo country, and consequently bad Indian country. A note I penciled June 3d, 1864, runs thus: "Our route since leaving Larned has been mostly along the north bank of the Arkansaw. Queer river that—a great ditch, chock full of grassy islets, stretching through the treeless prairie like a spotted snake, some seasons so dry you can't wet your foot in it for miles, and have to dig for a drink, sometimes a raging flood 200 yards wide. Traveling without military escort is risky. The Cheyennes are on the rampage; Comanches and Kiowas too." On the 6th, nearing Fort Lyon, we passed an Indian camp; "it was a band of Arapahoes, at war with the Cheyennes."

[II-21] No mileage for to-day. By Pike's map, camp is at an elbow of the river, which denotes that curve the Arkansaw makes in passing from Ford into Kiowa Co. There is no place to name in this vicinity, and the best maps, on a scale of 2 m. to the inch, do not give any island hereabouts. We will allow Pike 16 m., and set camp in Ford Co., just over the border of Kiowa.

[II-22] To "crease" a horse is to hit him with a bullet somewhere along the nape of the neck, close enough to the cervical vertebræ to stun him by the shock to the spinal cord, or to the ligamentum nuchæ, yet not to inflict permanent injury. When this is nicely done the horse falls as if killed, and is roped before he recovers. But it takes a very good shot, like "driving the nail," "snuffing the candle," "barking the squirrel," and other feats of skill which our backwoodsmen used to practice.

[II-23] Since he left Great Bend, Pike has had hilly country continuously on his left, with only a very narrow river-bottom on that side, in comparison with the breadth of the low-lying land on the W. or N. In fact, it is this series of countless thousands of hills and hillocks which causes the deflection of the river northward, thus making the "great bend." The place where the change occurs, and where Pike camps, is at Ford, a town in the county of that name, on the S. bank of the Arkansaw, or rather on the E. and S. bank of Mulberry cr., a stream from the S. W., which winds around the town on the W. and N., and falls into the Arkansaw a mile or so lower down. A branch of the Chic., Kas., and Neb. R. R. runs through Ford from Bucklin to Dodge City, Ensign, and Montezuma.

[II-24] Taking Pike past the site of old Fort Dodge and of present Dodge City, nearly to the boundary between Ford and Gray cos.—say halfway from Dodge to Cimarron, and thus about opp. Howell station of the A., T., and S. F. R. R. Dodge started on the N. bank, but has overgrown the river, and is now built up on both sides, with two bridges across. Dodge is 17 m. by rail above Ford, and almost exactly on the 100th meridian—probably some of the houses are built on each side of this line of longitude. At or near Dodge were the long-noted "Caches," of which most of the early travelers speak, but which seem to have been latterly lost sight of. I cannot locate the exact spot, but it ought to be easily recoverable by those who have the data I happen to lack. The place used to be spoken of as near the meridian just said—though that does not help us at all, as the maps of those days were mostly 30´ out of the way in longitudes. Thus, even Gunnison and Beckwith's route-map of 1853 runs the line E. of the mouth of Mulberry cr. where Ford now stands, and thus about 99° 40´. Gregg's is much closer than this, though it is on a much smaller scale; his 100th line runs midway betwixt the mouth of Mulberry cr. and the "Caches." Wislizenus' route-map, accompanying his report to Congress (Senate Misc. Doc. No. 26, 30th Congr., 1st Sess., pub. 1848) is closer still; for the "Caches" are marked scarcely W. of 100°. Wislizenus gives us another clew, as he marks "Fort Mann" at the "Caches." The "Caches" were also about the place where the dry cut-off, described in note¹⁸ above, reached the Arkansaw—in short, everything points to the immediate vicinity of Fort Dodge as the place where these caches were located. "The history of the origin of these 'Caches' may be of sufficient interest to merit a brief recital," as Gregg says, Comm. Pra. I. 1844, p. 67, where, and on p. 19, we have the account. In 1812 was fitted out the first expedition which attempted to reach Santa Fé by following the account of Pike's journey now before us. This consisted of about a dozen men, among them two named Beard and Chambers, who had succeeded in reaching Santa Fé with the others, and had returned to the United States in 1822 (Chambers had done so by way of the Canadian r.). These two interested some St. Louis capitalists to join an enterprise in the Santa Fé trade, and then undertook to return to Santa Fé in the fall of 1822 with a small party and an assortment of merchandise. "Reaching the Arkansas late in the season, they were overtaken by a heavy snowstorm, and driven to take shelter on a large island. A rigorous winter ensued, which forced them to remain pent up in that place for three long months. During this time the greater part of their animals perished; so that, when the spring began to open, they were unable to continue their journey with their goods. In this emergency they made a cache some distance above, on the north side of the river, where they stowed away most of their merchandize. From thence they proceeded to Taos, where they procured mules, and returned to get their hidden property. Few travelers pass this way without visiting these mossy pits, many of which remain partly unfilled to the present day."

[II-25] Alluding to the terrible defeat of General Arthur St. Clair's army by Indians on a branch of the Wabash r., in present Darke Co., Ohio, Nov. 4th, 1791. This was the most disastrous battle ever lost by the whites to the Indians, surpassing Braddock's defeat on the Monongahela in 1755. On Dec. 25th, 1793, General Anthony Wayne, who had become commander-in-chief in 1792, and taken command of the Army of the West, sent a detachment of soldiers to take possession of the field where General St. Clair had been defeated, built a fort there, and named the place Recovery, because it was then first recovered from the Indians, who had retained possession after the disaster above named. June 29th, 1794, General Wayne sent troops with supplies to Fort Recovery from Greenville, where he was then stationed. The detachment reached the fort and deposited its supplies in safety, but was immediately attacked, and the fort itself was invested by Indians, assisted by whites from Canada. The battle raged June 30th and July 1st, when the assailants were repulsed, not without great loss on our side. Among those who fell was the gallant McMahon, who had commanded the expedition to Fort Recovery. For further information see: Howe's Hist. Coll. Ohio, under head of Darke Co.; Burnet's Notes of the N. W. Terr., chap. vii; Albach's Annals of the West, p. 642. Present Fort Recovery is a village in Mercer Co. O., on a branch of the Wabash r., close to the Indiana State line.

[II-26] Camp past Cimarron and Ingalls, but not far W. of the latter—5 m., perhaps. These are two towns on the N. bank, respectively 18 and 26 m. above Dodge City. Ingalls is the seat of Gray Co. The Amer. Sp. word cimarron means something wild, runaway, or unreclaimed, like maroon, and is applicable to an animal, a person, a place, etc. It designated the wild sheep of the Rocky Mountains (Ovis montana), gave name to one of the largest branches of the Arkansaw, and was early associated with a certain route from the Arkansaw to Santa Fé. The name of J. J. Ingalls was long prominent in Kansas politics and in national statesmanship, and at one time associated with the too-true statement that "purity in politics is an iridescent dream." Notwithstanding the injunction against truth-telling which the consequences of the scholarly senator's remark imply, I wish to speak as accurately as possible regarding the points at which the Cimarron route left the Arkansaw. There were two of these places, both of which Pike passes to-day, where the river was forded, and the road thus crossed from the N. to the S. bank. These became known as the Lower and Upper Crossings of the Arkansaw; they were 8 m. apart; the lower one was 18 m. and the upper one 26 m. above Fort Atkinson; they thus correspond to the positions of Cimarron and Ingalls, respectively. The river is now bridged at each town. The Lower Crossing was the earlier one, most used by the traders from 1834 till the closing of the Mexican ports in 1843; after the war the Upper Crossing seems to have been generally chosen. Thus, we find Gunnison and Beckwith saying in 1853, P. R. R. Rep. II. 1855, p. 26: "Seventeen miles from the fort [Atkinson] there is a ford, sometimes used by the trains and parties going to and from New Mexico by the Cimmaron [sic] route; but the principal ford for that route is 8 m. above this." Writing of 1846, Dr. Wislizenus speaks of moving "about 20 miles" up the Arkansaw from the Caches, and arriving "at the usual fording place," i. e., the lower one. "This track," says Gregg, Comm. Pra. I. 1844, p. 311, "which has since remained permanent, was made in the year 1834. Owing to continuous rains during the passage of the caravan of that year a plain trail was then cut in the softened turf, on the most direct route across this arid desert, leaving the Arkansas about 20 miles above the 'Caches.' This has ever since been the regular route of the caravans; and thus a recurrence of those distressing sufferings from thirst, so frequently experienced by early travellers in that unhospitable region, has been prevented." The first camp S. of the Arkansaw was usually made in the vicinity of the Sand Hills, at a place called the Battleground after 1843, in which year the defeat of the Mexicans by the Texans under Colonel Snively occurred on that spot; it was some 12-15 m. from the river. The roads from the two fords came together at no great distance from the Arkansaw (perhaps in the vicinity of Ulysses, seat of Grant Co.); having thus headed the Crooked Creek branch of Cimarron, the road crossed Sandy cr. not far above its confluence with the Cimarron, and so reached that river.

[II-27] Past Pierceville, a village and station on the A., T., and S. F. R. R., just over the line between Gray and Finney cos.; camp 3 or 4 m. short of Garden City, seat of the latter county.

[II-28] Past Garden City and Sherlock; camp on or near the boundary between Finney and Kearney cos., in the vicinity of Deerfield, a place on the railroad. Most of the older maps mark hereabout the large island in the Arkansaw called Chouteau's, somewhat W. of the 101st meridian, and apparently near Deerfield.

[II-29] Vicinity of Harland, seat of Kearney Co. In saying that the Spanish road had been "on the outside" of the party, Pike gives us to understand that it had run along to his left, a little further from the river, though since the 30th of Oct. he had been also traveling on the S. side of the Arkansaw, having that river on his right. Nevertheless, the map marks the two trails as identical, the Spanish camps alternating with the American all along. There has been little to note along this stretch of the river, where no stream of any consequence falls in on either side. Pike here remarks a change, in the beginning of hilly country; extensive sand-hills are skirting the river on the S., in Kearney Co., and thence into Hamilton.

[II-30] Vicinity of Syracuse, seat of Hamilton Co.

[II-31] Last day's journey in Kansas, passing from Hamilton Co., over the inter-State line, into Prowers Co., Colorado. Pike's mileages along the whole course from Great Bend are remarkably close. I designedly ran them off day by day, without any checking by known positions, to see when he would strike the inter-State line, about 5 m. beyond which is the first identifiable named stream; expecting then to hark back, much as usual, and make the requisite adjustments of camps by proportionate lengths of each. But I find no occasion for this; his own mileages fix his camp of the 11th as nearly as possible on the line, and we have three identifiable streams in the course of his march on the 12th. To-day's camp is between Coolidge, Hamilton Co., Kas., and Hollys, Prowers Co., Col., 2 m. W. of the former, 4 m. E. of the latter, in lat. 38° 02´ N., long. 102° 02´ W.

[II-32] In Colorado Pike first comes opp. Hollys, a village on the N. bank and station of the A., T., and S. F. R. R. Below this are some small runs on the N., among them one called Cheyenne cr.; and Wild Horse cr. falls in on that side a mile above Hollys. He then crosses Two Butte cr., a much larger stream, from the S., arising in Las Animas Co. about the elevations from which it takes name, running through the N. W. corner of Baca and traversing Prowers to fall in a mile above the mouth of Wild Horse cr., opposite the large island there. Continuing, Pike crosses Granada cr., from the S., which falls in where the railroad crosses the Arkansaw and runs into the station named Adana. If he held straight on the best road, keeping to the left of the extensive bottoms along here, he went through the present sites of Granada, a village 4 m. W. of Adana, on Wolf cr., and of Manville, a station 2 m. further along. Camp was set about halfway between Manville and Carlton, a place 4 m. beyond. Several runs or washes make in along here on each side, but seldom carry as much water as the ditches which have been brought from the Arkansaw through and by Granada. Pike charts Two Butte cr., and one that answers either to Granada or Wolf cr.: notice the pair he lays down, S., with the legend "Cotton Wood becomes frequent" lettered across Two Butte cr.

The Wild Horse cr. above mentioned appears on Gregg's map by the name of "Lit. Sand Cr."

[II-33] No mileage to-day; and the omission is not easily supplied. On the 15th Pike camps at the mouth of Purgatory r., and it took him 34 m. by his reckoning to get there from his camp of the 13th. Therefore, camp of the 13th was about 12 m. from that of the 12th, and thus within a mile or two of Lamar. I shall so suppose it to have been. This sets Pike past the "point of red rocks and one large [Big Sandy] creek," which he speaks of as having passed on the 14th, but it agrees with the map, which sets a camp-mark for the 13th past Big Sandy cr. There is evidently a confusion of the record of the 13th and 14th, perhaps in the flurry of the Indian sign; all things considered, I shall set camp of the 13th, hypothetically, 2 m. short of Lamar: and that of the 14th at the station Prowers, 10 m. further; whence it is about 24 m. for the 15th to Purgatory r. The points passed on the 13th and 14th are most conveniently discussed together: see next note. The site of Fort Aubrey (named for or by F. X. Aubrey?), on the N. bank, was probably passed on the 13th.

[II-34] From his camp of the 12th Pike passes the village and station Carlton, opposite which the small Cottonwood cr. falls in from the N., and proceeds to his own "large creek" and "point of rocks." This stream is Big Sandy cr., from the N.; Pike lays it down very well. It is quite a river or river-bed, which when it runs drains from the high country known as the Arkansaw Divide, sc. between Arkansan and Missourian waters, in El Paso, Elbert, and Lincoln cos. The stream further traverses Cheyenne and Kiowa cos., and seeks the Arkansaw in Prowers Co., 2 or 3 m. below the point of rocks Pike notices. This is a place where a bold headland abuts against the river on the south, rising rapidly from 3,575 to more than 3,800—that is, some 300 feet above the general level of the river bottom. A run known as Clay cr. comes around the bluff on the W. The next above is Willow cr., S., on which Lamar stands between irrigating ditches derived from the Arkansaw, and the next above is Dry cr., S., halfway between Lamar and Prowers station. Here is camp of the 14th, just over the border of Prowers, in Bent Co. Pike's map legends "Broken with small Ravines & Creeks" on the country passed over.

[II-35] This statement conflicts with Pike's map, which lays down only one stream between the two camp-marks that stand for the 14th and 15th. But the text is right, and both these camp-marks are misplaced. One belongs just below Mud cr., and the other at Purgatory r., where there is no sign of one, though this is the most exactly locatable station since we left Great Bend. Pike's "two deep creeks" are Mud and Caddoa; his "many points of rocks" appear on any good topographical map. There is a series of such between Prowers and Mud cr., on the S., opposite which Graveyard cr. falls in, N. Two very notable points of rocks, a mile apart, are separated by Caddoa cr.; and Limestone cr. falls in from the N., 2 or 3 m. below these. These bluffs extend to the village of Caddoa, 2 m. up, in a bottom left by their recession from the river, before they again close in on the river in two bold headlands, 1 or 2 m. above Caddoa. The country on the N., across the river, is also bluffy for several miles along here. The elevations close to the river are 3,800 to 3,900 feet, and higher further back on both sides. Above the Caddoan bluffs a creek which Pike charts falls on the S. This is lettered Blue cr. on late G. L. O. maps, and Rule cr. on those of Hayden and Powell. Caddoa cr. heads about the N. W. corner of Baca Co., and takes a northerly course to the Arkansaw; Blue or Rule cr. is the larger one of the two; some of its affluents are near those of Caddoa and upper reaches of Two Butte cr., about Shell Rock cañon in Baca Co., but its real source is further south in Las Animas Co., where Johnny cr. and others head. Its course is northerly, but with an eastward trend, about parallel with Purgatory r. About an hour before Pike reached this large river he passed opposite the place where Fort Lyon was later built, on the bluff around which the Arkansaw there sweeps closely. In 1864 Lyon was the first inhabited place on the Arkansaw west of Larned, though there had been trading-posts or certain other temporary dwellings at various points, especially at the upper end of the Big Timbers, say 12 m. E. of Fort Lyon. These were a large body of cottonwoods extending thence several miles down the river on its N. side, and formed a noted resort of various Indian tribes. Hence the woods became well known to travelers along the Arkansaw, whose itineraries almost always speak of the "Big Timbers" as they approach the Purgatory on their way to Bent's fort. Pike's text of the 13th is no doubt the earliest indication of these woods.

Gregg's map lays down three large creeks from the S. between his Big Sand cr. and Purgatory r. The first of these is called Mulberry; the other two are nameless. The three appear to correspond to the Mud, Caddoa, and Blue creeks just described.

[II-36] The main chain of the Rocky mts., with Pike's Peak towering to the right: see L. and C., ed. 1893, p. 328. Pike has before him the Front range of the Rockies, northward, or to the right; and southward, or to the left, the Sangre de Cristo range. The sources of Arkansan waters are between these; while on the other side of the last named range are those of the Rio Grande. The "cheers to the Mexican mountains" were given at an alt. of 3,900 feet.

[II-37] Purgatory r., also called in English Picket-wire, in French Rivière Purgatoire, and in Spanish Rio Purgatorio and Rio de Las Animas, is charted by Pike as the "1st Fork," with the legend "Here the Mountains are first seen." This very large branch or fork of the Arkansaw heads in that southward continuation of the Sangre de Cristo range which is known as the Culebra range, about Trinchera, Culebra, and other peaks, where it connects with sources of the Rio Culebra, a tributary of the upper Rio Grande. Its own tributaries are very numerous and extensive. The main river runs N. E. from Las Animas Co., through the S. E. corner of Otero Co., and joins the Arkansaw in Bent Co., between the site of Fort Lyon and that of Las Animas, present county seat of Bent. Pike camped where the railroad now crosses; and his journey since the 12th has been practically along the present railroad line.

[II-38] It is certain that Pike was on Purgatory r. on the 15th, and certain that he did not reach his "Grand Forks" (present site of Pueblo, at junction of Fountain r.) till evening of the 23d. The distance between these points, along the river, is between 90 and 100 m. Pike's ostensible mileages are: for the 16th, 11½; 17th, 23½; 18th and 19th, none; 20th, 18; 21st, 21; 22d, 17; and 23d, 19; total, 110 m. We have, therefore, to reduce these mileages by about one day's journey. Observe, also, that only four camps are marked for the 16th-22d; there should be five, and with that for the 23d, six. Thus the text and map do not agree, and some error is evident, though what it is we have no means of deciding with confidence. I am inclined to think that the difficulty lies at the start from the "1st Fork" (Purgatory r.), when so much of the day was occupied in searching for the Spanish trail, and the "11½" m. assigned may have been little if any actual advance. If we proceed upon this supposition, there will be no trouble in adjusting mileages to bring in the missing camp by the 20th; after which all possible error is removed by the identifiable points. I shall, therefore, set camp of the 16th scarcely above the mouths of Adobe and Horse (formerly Dry) creeks, which fall in close together on the north, 7 and 8 m. above Purgatory r. Neither of these is noticed by Pike, though each is larger than some of the streams he charts. But they were across the river, and Pike had a bad case of Spanish trail on the brain, aggravated by anxiety about Indian sign.

[II-39] The mileage hardly requires any adjustment, from the position I have assigned for the 16th, to set camp of the 17th in the close vicinity of Rocky Ford, a village and station on the railroad, where Pike remains on the 18th and 19th. Rocky Ford is 5 m. above Timpas cr. (which he charts as the first stream from the left above his "1st Fork"). Timpas or Timpa is a large creek which heads in Las Animas Co. and runs N. N. E. into the Arkansaw at the upper point of that very large island above La Junta.

The most notable point passed to-day is the historic site of Bent's old fort, on the N. bank of the Arkansaw, 7 or 8 m. E. of the Timpas. It was a position of great consequence in the days of staging from Fort Leavenworth and other points on the Missouri to Taos, Santa Fé, and other New Mexican places. Most of the early itineraries, both commercial and military, speak of Bent's fort, and the advantages of this location for a post were more than once urged upon the attention of the War Department. It was on an emigrant road, in the heart of an Indian country overrun with various tribes; was a sort of focus for several widely divergent termini; was in the vicinity of good building material, and had plenty of fuel, grass, and water. Mr. Bent himself destroyed it in 1849, when he abandoned it for sufficient reasons; but some of the chimneys and adobe walls long stood to mark the spot. Thus it was Bent's "old" fort when I passed by, about 30 years ago. Gregg's, Wislizenus', and in fact most maps of the period mark the fort, some of them giving also a certain Fort William alongside it. The structure is described as "quite complete" by Lieut. J. W. Abert, who was here in August, 1846, at which time he met such noted plainsmen as Capt. Walker of California renown, Marcellus St. Vrain, and "Bill" Garey. Col. Price's regiment was here about the same time. The several columns of Gen. S. W. Kearny's Army of the West, which invaded and subjugated New Mexico and California, coming from Fort Leavenworth by the Arkansaw route, concentrated in camp a few miles below the fort, Aug. 1st, 1846. Kit Carson knew the place well, and Frémont found him not far away from here in July, 1843. A view of Bent's fort as it appeared in 1846 is given by John T. Hughes, in his admirable Doniphan's Expedition, 8vo, Cincinnati, 1847, p. 35. The old route into Santa Fé left the Arkansaw close by Bent's fort, went S. W. between Purgatory r. and Timpas cr., struck the latter at a place then as now called Iron Springs, and so on through the Raton mts., not very different from the way the A., T., and S. F. R. R. now takes. A glimpse at the kind of a road this used to be is had from the following extract from my diary: "Tuesday, June 7th, 1864. Bent's old fort. Cold ride in the rain from 3 a. m. to 5 p. m., when we brought up at the fort. Here was our crossing of the Arkansaw. Recent hard rains made the river unfordable; so we had to ferry ourselves over the surging tide in a frail skiff—ticklish business. However, we got safe across, with all our worldly goods—the latter nothing to speak of, and stood shivering while the ramshackled hack that met us on the other side was loaded and hitched up. This storm and the ferriage began a series of mishaps that reached to Fort Stanton in New Mexico, and made the driver swear that 'the grace of God had petered out on the other side of the Arkansaw.' Kept on to Iron Springs; road miry, pace snaily."

The name of Bent's Fort is preserved as that of a place nearly opposite (a little above) the present station Robinson, which latter is exactly on the boundary between Bent and Otero cos. Above this is La Junta, on the S., seat of the county. Several creeks fall in on the S. along here, the largest one of them named Crooked cr.

[II-40] Taking the Expedition just beyond the mouth of the Apishapa, Apishipa, or Apishpa r., to the present station Rockdale. This stream is charted by Pike; a camp-mark is set just above it, assuring us that the difficulty we had is already adjusted. It is a large river, or rather a long-bedded water-course (like many others which start well, but run out in the thirsty soil), heading about the Spanish Peaks, and reaching the Arkansaw at the foot of Apishapa bluffs (4,675 feet), between Rockdale and Catlin. Three miles off, across the Arkansaw, is the station Olney of the Mo. Pac. R. R., which here comes to the river. In old days a point opposite the mouth of the Apishapa was a good camp on the Cherokee trail to the gold-diggings on Cherry cr., with the Huerfano mountains and Spanish Peaks in sight.