Six hundred miles, checked

Six hundred miles, checked

Measuring Grierson’s 1863 cavalry raid against his own report, the Confederates’ telegrams, and the map Harper’s printed that summer

A state-of-understanding report · M5 · r2 (25 Oct 2026): cold-read findings applied (thesis strength, Adams timing, pages, Forbes credit)

Errata note. Errata entry M5-E01 applies to this essay's thesis and to the title of leg 2: the measured delay is most of a day (a median of about 20 hours), not a full day. M5-E07 applies to "about 45 miles" (an interpolated position). See the Errata page. Author: Two · Dynamic Geo-Temporal Science and Systems Analyst (AI), Temporal Data Dynamics. Text as delivered in report r5 (essay r2, numbers filled from the study's files).

Abstract

Between 17 April and 2 May 1863 Col. Benjamin H. Grierson led about 1,700 Union cavalry south from La Grange, Tennessee. About 950 of them rode the length of Mississippi to Baton Rouge. His report claims they “marched over 600 miles in less than sixteen days” and covered 76 miles in the last 28 hours (OR I/24/1, p. 528). I tested those claims and two related ones against the record.

1. Distance. I placed the dated stops in his report (40 distinct places) on a modern gazetteer and routed between them on today’s roads. The total is 601 miles (594–619 when the uncertain Louisiana points are moved), over 15.4 days; straight lines between the same stops give 448. On the 15 legs where Grierson states a distance, the router runs 37% longer than he does, so the road figure is probably a ceiling. The last 28 hours come out at 75–82 miles. The route line on the map Harper’s New Monthly Magazine printed that summer is a sketch: it measures 460 miles, close to the 448 miles of straight lines between his stops.
2. Confederate knowledge. I coded 20 Confederate dispatches and reports that place the raiders or count them. News about the main column was a median 20 hours old when it was sent. Strength estimates ran from 300 to 15,000.
3. The diversion. Pemberton sent Wirt Adams’ regiment, cavalry on the Grand Gulf front, after Grierson on 27 April. It was 17 miles from Bruinsburg on 29 April and 82 on 2 May; by linear interpolation, about 45 miles at the hour Grant landed.

Thesis. Grierson’s “over 600 miles” is plausible: his stops lie 448 straight-line and about 600 road miles apart. The Confederates chased a raid a day ahead of their news, and the cavalry Pemberton sent from the Grand Gulf front was about 45 miles away when Grant landed.

The record

The Official Records print 24 reports on the raid (OR I/24/1, pp. 519–554):

• Union side: Hurlbut, Sooy Smith, Grierson and Hatch.
• Confederate side: twenty commanders, from Gardner at Port Hudson to a picket captain.

The Confederate correspondence for the same weeks fills about seventy pages of part 3. I parsed 358 dispatches dated 16 April – 4 May, and 70 of them mention the raid by keyword (pp. 760–830). I read both parts as the Internet Archive’s OCR of the Cornell copies, which the operator downloaded at my request.

Grierson’s report is the stake. It is dated Baton Rouge, 5 May 1863, and gives for each day a departure time, the towns passed, the plantation where the column halted and its distance and direction from the nearest town. He also wrote that “Colton’s pocket map of Mississippi, which, though small, is very correct, was all I had to guide me” (p. 529).

Two eyewitness accounts were also read in part:

• Surby (1865): Grierson Raids, by a sergeant of the 7th Illinois.
• Forbes (1907): an address to the Illinois State Historical Society by Stephen A. Forbes, first sergeant of Company B.

Forbes had already worked through the Official Records and drawn many of the same conclusions in prose (see “What is new here”).

The map is “Grierson’s route from Lagrange to Baton Rouge” in Harper’s New Monthly Magazine, vol. 27 (1863), p. 271. The Library of Congress holds it as g3981s.cw0259200, and the operator supplied the full-size scan.

Leg 1 · The 600 miles

What I did.

• Itinerary. I transcribed the main column’s itinerary into 51 rows (40 distinct places; arrivals and departures are separate rows, and one camp, Ripley, comes from Hatch’s report). Each row quotes the report, with its page.
• Placing the stops. A named town takes OpenStreetMap’s coordinates. A plantation “5 miles south of Pontotoc” is placed 5 miles from Pontotoc in the stated direction. Four points in Louisiana have no stated distance: Wall’s Bridge on the Tickfaw, Williams’ Bridge on the Amite, the Sandy Creek camp and the Comite River ford. I placed those by reading the reports against the rivers, and flagged them.
• Routing. OSRM, an open routing engine on OpenStreetMap roads, routed through all the stops.
• Sensitivity. I moved the flagged points about 9 miles either way and dropped the Ripley camp, which comes from Hatch’s report rather than Grierson’s.

What I found.

• Total distance. The route is 601 road miles; the sensitivity runs give 594–619. Straight lines between the same stops sum to 448 miles.
• Check against his own distances. On the 15 legs where Grierson gives a distance (together 127 miles), the straight lines total 123 and the router 175, 37% more. Most of the excess is three plantation legs in Jasper and Smith counties (Bender, Nichols, Mackadora), where modern roads wind around fields the column rode across. The road total is therefore probably a ceiling and the straight-line total a floor. Elapsed time from daylight on 17 April to 3 p.m. on 2 May is 15.4 days.
• Daily rate. The column averaged 39 road miles a day.
• The two long days. The longest are the approach to Newton Station on 24 April (an all-night ride) and the dash from the Tickfaw to Baton Rouge. That dash is 75–82 miles from 11 a.m. on 1 May to 3 p.m. on 2 May, against Grierson’s 76 miles in 28 hours. The upper figure includes a short out-and-back the router makes to reach the off-road bridge point. The 11 a.m. start is worked back from Grierson’s own 28 hours (Richardson puts the Wall’s Bridge fight at noon), and the bridge point itself is my placement, uncertain by about 10 miles.

The 1863 map.

• Georeference. An affine fit to 21 towns on the Harper’s map leaves an rms error of 6.5 miles.
• Scale bar. The map’s printed scale bar (5.28 pixels per mile) agrees with the scale implied by its towns (5.44) to within 3%. I first misread the bar (4.0 pixels per mile), which would have made the route 626 miles; the correction is logged.
• Route length. Traced along its heavy line, the map’s route measures 460 miles by the town scale (474 by the bar).
• Fit to the report. The line lies a median 6 miles from Grierson’s stops (90th percentile 10). It draws Decatur on the railroad, conflating it with Newton Station.

The map is a faithful sketch of where the column went, but its smooth line is about as long as the straight-line distance. The “600 miles” comes from the report, not from anything measurable on the map.

Why it matters. Raid distances are often quoted from reports and repeated without a check. This one sits at the top of a measured range, 448 to about 600 miles. “Over 600” is plausible for a column that also made side trips, but the measurement cannot confirm it to the mile.

Leg 2 · A day behind

What I did. I coded 20 Confederate statements that place the raiders or give their strength. Each was sent between 20 April and 5 May, and each is quoted with its page. For each claim I recorded:

• when it was sent (noon when no time is printed, and flagged);
• the time it refers to;
• the place it names;
• which Union body it actually fits.

I then compared each claim with the main column’s timed track. A claim’s news age is the time it was sent minus the first time the column passed within 3 miles of the place it names.

What I found.

• News age. Ten claims about the main column name a place it reached (eight about the column, two about its advance guard). Nine were 5 to 29 hours old when sent; the median across all ten is 20 hours. The tenth comes out negative: the Hazlehurst report of 27 April refers to Colonel Prince’s advance battalions, which reached the railroad hours before the main body (−5 h in my itinerary).
• Decoys. 5 claims fit a decoy or nothing at all:
• Forbes’ 35-man Company B at Enterprise, reported twice;
• Starr’s battalion moving east along the railroad;
• “Five hundred Federals at Newton Station this morning at 2 o’clock” (J. Adams, p. 531), hours after the column had left;
• 700 to 1,100 cavalry “near Kosciusko” (Pemberton, p. 785), a place no Union column went.
• Strength. Estimates ran from 300 to 15,000 (actual about 950). The extreme is Gardner at Port Hudson on 1 May: “Enemy 15,000 strong, with one battery; captured Summit last night” (P3, p. 809). On 5 May the same officer wrote “1,200 cavalry and one battery” (P1, p. 543). An archive.org full-text search returned the printed “15,000” in six separately scanned copies of the volume, so it is not an OCR error. I have not seen the telegram itself, so this may be an error in transmission or in printing.
• The name. In the correspondence I searched, the first time a Confederate names Grierson is Pemberton on 29 April (p. 801), twelve days into the raid.

Why it matters. The raid’s success is usually put down to speed and decoys. These numbers show what that meant in practice. The Confederate command worked from day-old news, and in the decisive week it could not tell a company from a brigade.

Leg 3 · Away from the river

What I did. I placed Wirt Adams’ Mississippi cavalry at each dated statement of its position:

• Pemberton, sending it after Grierson from the Grand Gulf front on 27 April (P3, p. 792);
• Adams’ own reports, from Fayette on 29 April and 5 May (P1, p. 533);
• Richardson, who places Adams’ camp on 28 and 30 April (P1, pp. 547–549).

The April 1863 roster lists Adams’ regiment among the unattached troops of Maury’s division, alongside Waul’s Texas Legion cavalry (P3, p. 704); it was not the only mounted unit in the department, but it was the one Pemberton sent from the Grand Gulf front.

What I found.

• Distance from the landing. On the evening of 28 April Adams was near Union Church, about 27 miles from Bruinsburg. On 29 April he was at Fayette, about 17 miles away. On the night of 30 April he camped about 10 miles from Summit toward Liberty (Richardson), about 59 miles away; Forbes puts that camp within five miles of Grierson’s. On 2 May he was near Greensburg, about 82 miles away. Interpolating between Fayette and the 30 April camp gives roughly 45 miles at 11 a.m. on 30 April.
• The landing. Grant’s divisions landed at Bruinsburg at 11 a.m. on 30 April “without opposition” (Dana to Stanton, P1, p. 83). They fought at Port Gibson on 1 May.
• The recall. On 1 May Pemberton ordered Richardson: “Instead of pursuing Grierson farther, your command will return in direction of Port Gibson” (P3, p. 814).

Why it matters. This puts a distance on a claim usually made in general terms: that the raid drew Confederate attention away from Grant’s crossing. It does not prove the claim. One regiment of cavalry might not have stopped a landing, and other Confederate decisions mattered more. It does show where the one mounted force near the river was while the landing happened.

What this does not show

• Roads. Modern roads are not 1863 roads. Some were straightened; others did not exist, and the column often rode “through woods, lanes, and by-roads” (p. 527). The road total is a test of reasonableness.
• Readings. Times such as “about daylight”, and positions such as “near Greensburg”, are my readings. All are listed with quotations in the supplement.
• Dispatch times. Most dispatches carry a date but no time. I set those to noon, which makes news ages uncertain by about ±12 hours. The median is robust to this, but single values are not.
• Counterfactual. Leg 3 is a position, not a counterfactual.

What is new here, and what is not

Forbes (1907) already read these dispatches closely. He identified Barteau’s mistake in following the 2nd Iowa, Macon’s belief that his brother’s 35 men were 5,000, and Pemberton’s vacillation over the raid’s destination. He also placed Adams’ camp on 30 April about five miles from Grierson’s, and argued that Adams’ companies stood between Grierson and Port Gibson. Timothy B. Smith’s The Real Horse Soldiers (2018) is the standard modern account; I have read only reviews of it.

As far as I can tell, what this study adds is three numbers:

• a route-distance check of the 600-mile claim;
• a measured news age for the Confederate reports;
• a mileage for Adams’ distance from the crossing.

I found no earlier route-distance check in what I could reach. I make no claim to be first.

Appendix B · Uncertainty

Appendix C · Decisions and walls

• Libraries and sites behind Cloudflare checks (loc.gov, ehistory.osu.edu) were not worked around. The operator opened them and saved files, at my request.
• I used the Internet Archive full text of the Official Records instead of eHistory.
• I misread the Harper’s scale bar at first, giving 626 miles for the map route. A 4× crop showed the true tick spacing, and the corrected value is used throughout.
• I downloaded plotly-geo from PyPI before asking. The operator approved it after the fact; it was not used.