A cool account

67th Tigers

1st Lieutenant
Nov 10, 2006
I am also genuinely interested. Those who analysed the problem at the time, such as the historian of the 6th US, understood the issue; the US Cavalry lacked carbines and so couldn't operate without infantry support. By 1863 they have sufficient carbines and numbers to operate without infantry support.

As to the cavalry during Stuart's Raid, it was mostly under Stoneman north of the Chickahominy. South of the Chickahominy was only the 3rd Pa Cav, a squadron or battalion of the 8th Illinois Cav, the 3rd Bn of the 6th NY and that attached to McClellan's HQ.

Stoneman's command consisted of a line of pickets consisting of the 8th Illinois from the Chickahominy to Pole Green Church, with the 1st NY Cavalry encamped with the base of the 8th Illinois as a reserve for that part of the line. At Pole Green Church was a combat outpost of the 5th US Cavalry (Royall's battalion, 2 sqns), with a coy pushed out to watch Totopotomoy Creek there. The 8th Pennsylvania Cavalry picketed the length of the creek to the right of the Church. The rest of the reserve (1st, 5th(-) and 6th US and the 6th Pa Cav) were concentrated in a camp near Old Cold Harbor.

Stuart punched through Royall's battalion of the 5th US Cavalry, and the int Royall sent back was that the enemy had a division of infantry with them. Cooke immediately reacted by sending the 5th (-) and 6th US to gain contact with this force, and sending for a brigade of infantry. The 6th Pa Cav and 1st US Cav followed. The next day the 3rd Pa Cav moved out to try and head Stuart off.

The main body of the pursuing force was concentrated, and had enough combat power. The problem was finding Stuart, which necessitated dispersing patrols. One of the scouting patrols did intercept Stuart, but was overrun and captured.

There was no really better way to organise the available cavalry that I can see.


Feb 18, 2017
As to the cavalry during Stuart's Raid, it was mostly under Stoneman north of the Chickahominy. South of the Chickahominy was only the 3rd Pa Cav, a squadron or battalion of the 8th Illinois Cav, the 3rd Bn of the 6th NY and that attached to McClellan's HQ.

Stoneman's command consisted of a line of pickets consisting of the 8th Illinois from the Chickahominy to Pole Green Church, with the 1st NY Cavalry encamped with the base of the 8th Illinois as a reserve for that part of the line. At Pole Green Church was a combat outpost of the 5th US Cavalry (Royall's battalion, 2 sqns), with a coy pushed out to watch Totopotomoy Creek there. The 8th Pennsylvania Cavalry picketed the length of the creek to the right of the Church. The rest of the reserve (1st, 5th(-) and 6th US and the 6th Pa Cav) were concentrated in a camp near Old Cold Harbor.
So I make that:

Forming the screening line = 36-38 coys

8 or 10 coys of 8th IL
4 coys of 5th US
1st NY (12 coys)
8th PA (12 coys)

Reserve north of the Chickahominy = 34 coys

1st US (4 coys)
5th US (-) (6 coys)
6th US (12 coys)
6th PA (12 coys)

South of the Chickahominy = 18-20 coys

2 or 4 coys of the 8th IL
3rd PA (12 coys)
3/6th NY (4 coys)

HQ attached and not otherwise mentioned = 12 coys

2nd US (7 coys)
McClellan dragoons (2 coys)
Oneida coy (1 coy)
Det. 4th US (2 coys)

And 5 coys of the 11th PA may not have arrived yet, while the 4th PA definitely hadn't.

So the flank on which Stuart's raid took place was by far the more heavily picketed, with McClellan placing approximately 70 coys there out of a total available of about 102 (split roughly halfway between the picket line and the reserve). At this point there is no benefit to be gained by cutting the "south of the Chickahominy" and "HQ attachment" roles, not unless a crisis has already started to happen, and until Stuart moves there isn't a crisis going on; the only meaningful place from which you can reinforce the reserve is to strip the picket line, and vice versa.

If the whole force is on the picket line then there's no quick reaction force; if almost the whole force is in the QRF then there's nothing to stop even a raid in battalion strength, because the picket line has to cover a frontage totalling 10+ miles. As it was Stuart used, what, the equivalent of two fresh regiments to break through the screen...


67th Tigers

1st Lieutenant
Nov 10, 2006
That's reasonable. A squadron of the 6th US were out on patrol that day, and came close to unwittingly intercepting Stuart. There may have been another outpost manned by the 6th US east of Royall's. The 1st NY appear to have pickets along the Chickahominy along "no-mans land".

67th Tigers

1st Lieutenant
Nov 10, 2006
At 2300, McClellan received this report of the dispositions north of the river:

Received: 11.00 P.M.
From: Porter’s Headquarters
To: Gen. Marcy

The following is the disposition of the command.

Stoneman scouts [1] all roads leading into Mechanicsville road, in rear of Cooke, and also nearly to Peake’s station [2] & connects with Cooke.

Cooke is within a mile and a half of Old Church with (1) battery, 5th​ & 6th​ Cavalry & 1st​ Cavalry and Warren’s brigade minus 4 companies 5th​ & 6th​ Cavalry cut up or captured at Old Church & 2 companies have gone with Emory.

Emory with 4 squadrons Rush’s cavalry has gone to Tunstall’s Station. One (1) squadron from Cooke to join him – Cooke has orders to scout everywhere in his front and report after. If a superior enemy attacks him, he will ascertain where the enemy [is] so that when Sykes with his division and (3) regiments from Morell & 1 battery, the enemy may be whipped.

Sykes will move at them (3) o’clock and will go on ‘til he is informed by Cooke that the enemy in force has gone. If notified before (3) he will not move.

Infantry and Cavalry are on the road from Cold Harbor to Old Church. Emory gone to Tunstall’s Station. Meadow’s Bridge [and] Mechanicsville Roads will [be] guarded & patrolled.

Each bridge is guarded on this side as well as the other.

All troops are held ready to move at a moments notice. Details for labor made for tomorrow which will be interrupted only by a fight.

F.J. Porter

[1] Stoneman commanded the cavalry in general, and was commanding the 8th​ Illinois, 1st​ NY etc. as a brigade.
[2] Peake’s Station is the stop before Hanover Court House, about 4 miles south of it.

Edit: originally transcribed McClellan received version, but the sent version is also in the papers and was used to fill in the unreadable parts.
Last edited:


Jun 2, 2011
Washington, D.C.
Lt. Edward H. Leib was from Pottsville, Pennsylvania. He was commissioned into the 5th U.S. Cavalry in 1861 despite having no formal military training. He made the Regular Army his career and retired as a full colonel in the 1890's. The photo below has his middle initial wrong.

View attachment 379863

In doing some research on his role in the Battle of Monocacy, I found this account of the March 17, 1863 Battle of Kelly's Ford that I had never seen before. It's a good one. It was published in his hometown newspaper, the Pottsville Miners Journal:

Camp Near Falmouth, March 20, 1863.

I suppose you have heard of the cavalry, and of the success of the last trip we were on over the Rappahannock. We crossed at Kelley's Ford; had quite a fight with the rebels, and have taught them one thing, that we can whip them in a fair stand-up fight. We left camp for our trip on the 16th, and arrived at Morrisville at dusk. There we camped over night, and at two in the morning we started for Kelley's Ford, and there met the enemy. We had quite a time in crossing, but we were determined to cross, and we did. I am sorry, however, to state that we lost some good men while effecting the passage. We took about twenty-five prisoners and killed several of the rebels. We then, after getting the artillery over safely, moved on the road for Culpepper Court House; but we had not gone far before our cavalry came upon General Lee's brigade with himself at its head. They made a charge, but our men met them splendidly and drove them back. But they were not satisfied, and soon came over on our right flank. I must here state they made a grand mistake. The Fifth and First Cavalry were there, and your humble little friend had the honor of commanding the Fifth on the occasion. I was ordered to charge, which I did, leading the gallant regiment. We drove them, and I suppose they will admit that they were never driven so before. We kept it up until they got out of sight, and we were ordered back by the General, or rather Captain Reno, who commanded the brigade. Captain Baker had command of the First Cavalry. I was then ordered to move up in line of battle with the regiment, which I did through the thick woods and marshy ground, into a clearing. It was hot work to get there, but we made them leave, and obeyed our orders to drive them. As soon as we arrived about two hundred yards in the opening, they opened one whole battery on my command. It was rather a hot place, but the men stood it like Spartans, and held their ground until ordered to fall back, which was done in splendid style. We again foiled the enemy under the hottest fire I ever saw. The men were a little confused, but did not break or straggle. When the enemy saw our line moving back, their cavalry made a charge down the road. We could see that they meant to do some tall charging, but we moved up to meet them with drawn saber, and they turned and tied. They do not like our cold steel. They here broke and ran up to their entrenchments, scattering in every direction. It was the finest little fight I ever saw, and the old Fifth had the work to do. The regiment had the advance after crossing the ford on the other side, and in conjunction with the First United States Cavalry, had the rear guard in crossing the ford. On this side of the river I had the rear guard back to Morrisville.

I would not have missed the fight for a great deal, and hope soon to again show the country that we can whip the rebel cavalry every day in the week. The army is now in fine spirits, and our cavalry fight is all the talk in camp. The cavalry are for the present the tigers of this army, and hope soon again to meet the rebel cavalry.

Yours, E. H. L.
This is a wonderful account. Thank you

James N.

Forum Host
Annual Winner
Featured Book Reviewer
Asst. Regtl. Quartermaster Antietam 2021
Feb 23, 2013
East Texas
I’ve encountered Marcus Reno in 1864 when he was a staff officer

As I mentioned previously, Reno shows up as a member of a staff in Jeffery Wert's Mosby's Rangers in April, 1865 during the negotiations for the surrender of Mosby's command.


Feb 18, 2017
I think one of the ways to conceptualize cavalry raids and anti-cavalry stuff is that it's about time, motion, preparation, and the available routes.

Expanding on my map from above, once Stuart has broken through the screen with a force roughly equivalent to two regiments then what are the options?
If you try to run him down from behind, then you're basically betting on your own horses being better rested and/or faster than his - but he's launched this raid when he's ready for it, so he's if anything going to be the one moving faster in a stern chase.
You can instead try and get in his way, but you need quite a substantial force to be able to do that - he's got two regiments with him - and there are several possible routes he could take. Even assuming that all the cavalry leaving their starting point at Cold Harbor wouldn't just let Stuart take that route, I count at least three options Stuart has at just about all points.


That means it's a matter of trying to work out his route before he actually takes it, and "cut the chord" to get there faster than him - but before long that's not possible, because he's past Cold Harbor and moving away, and early on it's not possible because you don't know the route he's taking.

The easiest things to do are to defend your vital points (because you know where those are, and probably already have troops there for that matter) and to chase him down from behind, because that means Stuart's period behind your lines will be brief. But unless you have a massive superiority in resources you can't guarantee the defeat of a raid - as you need to be able to do one of these two things.

1) Have enough forces available to block every possible entry/exit point in such force that the cavalry raiders cannot defeat any one of them.
2) Have enough forces available in reserve to form a complete blocking cordon through your rear areas.

Without that, it's a matter of correct prediction, the enemy zigging when they should zag, and luck. Or the enemy making a significant error in the setup of the raid (like Streight's Raid, in which I undersand Streight's brigade had many of its men mounted on mules and which Forrest was able to catch up with.)

67th Tigers

1st Lieutenant
Nov 10, 2006
As McClellan himself wrote, he simply lacked the necessary numbers of cavalry:

"It was intended to assign at least one regiment of cavalry to
each infantry division so long as the division organization was
the highest, and, when army corps should be formed, to attach
a strong brigade of cavalry to its headquarters, leaving with the
division only enough for the necessary duty ; also to form a gene-
ral cavalry reserve. On the 15th of Oct. there were serving with
the Army of the Potomac, including General Banks's command,
one regiment and two companies of regular cavalry, and eleven
regiments of volunteer cavalry. When the army took the field
there were on its rolls four regiments and two companies of regu-
lar cavalry, eighteen regiments and five companies of volunteer
cavalry, besides four regiments yet unprovided with horses.

Of these there went to the Peninsula the regulars and four
regiments and five companies of volunteers, making eight regi-
ments and seven companies ; and there remained with Gen. Banks
and at Washington twenty-one regiments, besides the four unpro-
vided with horses. Circumstances beyond my control rendered
it impossible for me to carry out my views as to the cavalry, and
it was entirely against my wishes and judgment that I was left in
the field with so small a force of this arm. Of the field force one
regiment of regulars were necessarily employed on provost duty,
and two companies of regulars and one of volunteers at head-
quarters, leaving only three regiments of regulars and four regi-
ments and four companies of volunteers, certainly not over four
thousand men at most, to do all the cavalry and mounted orderly
duty for the army of eleven divisions — a force so ridiculously
insufficient, less than one-fourth of what it should have been, as
to render it strange that the enemy contented themselves with
riding around our lines only once on the Peninsula."


Feb 18, 2017
I did a quick calculation based on the returns for June 20 1862, June 30 1863, and April 30 1864.

June 1862:
I took the PFDE strengths of all the individual cavalry units with the corps and the reserve cavalry, and added 500 as an estimated amount spread between the QMG and provost, and compared it with the PFDE strength of the rest of the army (discounting Fort Monroe etc. as being more than 50 miles away)

June 1863:
I took the PFD strength of the cavalry division on this return (which is from the monthly return) and compared with the same category for the rest of the army.

April 30 1864:
I took the list PFD for the cavalry corps, and compared with the same category for the rest of the Army of the Potomac (not including 9th Corps).

Cav (PFD)​
Inf (PFD)​
Total (PFD)​
Cav %​
Jun 62​
PFDE but this is basically the same as PFD​
Jun 63​
Given as PFD​
Apr 64​
PFDE and is significantly reduced from PFD​
("inf" here means men in non-cavalry units)

Interestingly the regular cavalry on the March 1862 return doesn't embrace all the cavalry McClellan wanted to take, and if we assume that it's only about a third, for 2 brigades out of 6 (i.e. he wanted to take about 8,000 cavalry PFD) then the result is that he wanted to take:
Cav (PFD)​
Inf (PFD)​
Total (PFD)​
Cav %​
Peninsula, intended​
Mar 62​

The amount that originally reached the Peninsula in April (by which time Franklin's division had been returned) was about half that (7 regiments out of 14), so:

Cav (PFD)​
Inf (PFD)​
Total (PFD)​
Cav %​
Peninsula, by April​
Apr 62​
(this uses pre-Peninsula strengths of the units which arrived)

The difference is pretty striking, and actually presents a puzzle - in November 1862 the good cavalry performance in the Loudoun Valley campaign was still at about a 6% cavalry ratio:
Cav (PFD)​
Inf (PFD)​
Total (PFD)​
Cav %​
Nov 62​

(This uses PFD from the November 10 return, deducts "Defences Upper Potomac" from the overall total "Army in the field", and assumes the 1 RI, 6 PA and 6 NY attached directly to 3rd/6th/9th corps respectively are folded into their corps returns and are 500 men on average)

This means that any problems in cavalry performance in McClellan's army can be explained by the relative scarcity of cavalry, and in fact if anything the organization must have been superior to make up for the deficit.


Feb 18, 2017
In attempting to establish a minimum required cavalry allowance, I think there are two slightly different situations to cover, along with some commonalities.

Commonality - the escorts, provost and couriers.

This is something you need whether the army is in motion or static, and it's more an absolute rather than a relative amount. It's also not very large, and something which doesn't require full regiments, so it's a good place to put odds and sods of partial regiments like the Oneida Company.
It only takes a couple of regiments to provide for this.

In Motion - advance guard/rearguard

This is basically the cavalry who sweep ahead of the army and take control of bridges or fords, scout ahead and identify anything which would impede the advance of the main army, and so on. This is a cavalry job because it's a task for a fast moving unit.
This is basically something you need for the army itself, and you could really do with it for each formation acting as a corps (i.e. on a different route of advance) where relevant. Otherwise you end up with situations where a corps is advancing down a road that does exist on a map and leads to the right place, but they have no "eyes" to see if there's something in the way.

Commonality - scouts and screens

This is referring to the cavalry which reaches out to identify the enemy army's position, detect the approaching enemy, and prevent the enemy from doing the same to you. I would say that you might need more of it when static, at least if you're against an enemy defensive line, but then again you can just reassign the advance guard and rearguard. So effectively when static you can thicken up the scout/screen to see what your enemy might be trying to break the static situation.

In Motion - battlefield use

This applies when the army is fighting, but that usually means they're manoeuvering so I've marked this one as in-motion.

Static - raiding.

I've listed this one as static because it's generally a bad idea to send off most or all of your cavalry when you're manoeuvering - exceptions could exist, but often in the actual war when someone sent off a big raid during a manoeuvre period it resulted in a lack of cavalry at a critical time (cf. Stuart at Gettysburg, Hooker at Chancellorsville, Grant in the Overland).
This mode however requires a lot of cavalry, at least if you want to accomplish anything. So you need a pretty high cavalry fraction.

So by the looks of things you need more cavalry on the move than when static, unless you're going for a big raid. I'm unable to locate a reason you'd need more cav when static.

Meanwhile I did a rundown of Sheridan's field force in October 1864.

Sheridan in October, forces in the field, PFD:

Dept. of WV
12,186 inf
6,291 cav
3,307 art (counted with inf)

6th AC (all inf/art)

Det. 19th (all inf/art)


Cavalry Forces
7,078 cav
497 art (counted with cav)

Total cav:

Total field army:

Cav is 26.9% of field army

This is actually pretty similar (if not slightly higher than) the cavalry ratio at Waterloo. Sheridan by this count had an army about 75% of the size of Napoleon's Waterloo army, but the same amount of cavalry as Napoleon at Waterloo; it's vastly disproportionate to anyone else examined so far.

67th Tigers

1st Lieutenant
Nov 10, 2006
Indeed, it's hard to fathom just how well McClellan did with his available cavalry.

FWIW, I have Starr here. He bizarrely opines, "There is nothing to suggest that McClellan did not have a free hand to take with him as large a force of cavalry as he thought he needed, but he misused even the cavalry he did have. Had he organized his nine and a half regiments of cavalry [sic] into a three-brigade division, properly led at the brigade and division level and properly used... [It] is doubtful if a larger, better organized cavalry could have had a significant effect on the outcome."

Of course, literally everything about this is untrue. A simply perusal of the OR will find McClellan ordered the cavalry not taken away from him to join him on the Peninsula, and McDowell arrested Colonel Campbell of the 5th Pennsylvania for attempting to comply. McClellan tried literally everything to get cavalry regiments, but his requests were denied.

With this in mind, how should the cavalry have been organised according to Starr? By late April the cavalry on the Peninsula was:

Escort to Army HQ: 1 sqn, 4th US Cavalry
Scouts and Guides: elm. 2nd US Cavalry (7 weak coys organised into 2 sqns, in July Coys A, B & C were officially cadrised)

Commander, Cavalry (BG Stoneman) commands all formations below when concentrated

Attached to 2nd and 4th Corps: 3rd Bn, 6th NY Cavalry (2 sqns)
Attached to 3rd Corps: 3rd Pennsylvania Cavalry

Cavalry Division (BG Cooke)

1st Brigade (BG Emory)
5th US Cavalry
6th US Cavalry
6th Pennsylvania Cavalry

2nd Brigade
1st US Cavalry (4 sqns only)
8th Pennsylvania Cavalry
McClellan Dragoons (1 sqn)
Oneida Coy (0.5 sqns)

Here the only reasonable change one could make would be to swap the McClellan Dragoons for the 3rd Pennsylvania Cavalry.

As it was, the cavalry assigned to the corps turned out to be inadequate, and constant details had to be made to the reserve to the corps.

To this, McClellan had gained the grand total of 2 more regiments (8th Illinois and 1st New York) and a squadron of the 5th Pennsylvania (whose colonel was arrested for sending them) by 20th June. The ships carrying the 4th Pa Cav arrived off White House Landing on the 24th June. By this time the problem of having to provide a company to each division was solved by having the 2nd US Cavalry and McClellan Dragoons (a total of 9 coys, albeit weak ones in the case of the 2nd US) perform provost freeing the other regiments from having to make major detachments.

This gives 8 regts available to form a cavalry division; 3 regular (1st, 5th and 6th) and 5 volunteer (1st NY, 3rd Pa, 6th Pa, 8th Pa and 8th Illinois), although to provide the minimum level of escorts, several additional squadrons need taking from these as was necessary in the consolidation of 8th July).

I would argue, with the benefit of hindsight, that the 6th Pa Cav should have been removed from the cavalry division, converted to a corps of guides (as per 8th July) and assigned to provide corps escorts along with the 3/6th NY and the sqn of the 5th Pa. This would leave 7 regts, of which on would have to be posted south of the Chickahominy, leaving two triangular bdes. Arguably, Stoneman should have taken them as a division and made Cooke a brigadier.

However, the effect of this would likely be minimal, because in reality what ended up happening was that the cavalry north of the Chickahominy ended up operating as two bodies, under Stoneman and Cooke. Certainly there is no possibility of a "3 brigade division", unless you want very thin brigades.

However, the presence of the missing cavalry, the 5th Pennsylvania and 1st New Jersey, might have helped significantly. McClellan had lost the 2nd NY and 1st Pa Cav with the reassignment of 1st Corps, 4th NY Cav with their reassignment to Fremont and 3rd NY with their reassignment to Burnside, but those two regiments were legally McClellan's. Whether two more cavalry regiments would have made a difference is a question, but undoubtedly McClellan asked for them, and was denied.


Feb 18, 2017
So all the dribs and drabs amount to:

2nd US (7 weak coys)
McClellan dragoons (2 coys)
Oneida coy (1 coy)
4th US (2 coys)
3/6th NY (4 coys)
Sqn, 5th PA (2 coys)
Total 18 coys

The medium-strength units are:
1st US (8 coys)

And full-strength are:
5th US
6th US
6th PA
8th PA
3rd PA
8th IL
1st NY

And the problem is that you don't really have enough in the "dribs and drabs" to totally fulfil the "corps duties" (escort, provost etc). So you need to take at least one full-strength or nearly-full-strength regiment to do it.

I think there's an argument that you could have had:

Escorts etc.:
13 coys (2 coys per army or corps HQ, plus Oneida coy as specialists) taken from the McClellan Dragoons, 4th US, Oneida coy, and one volunteer unit

Provost: 2nd US

Floating assignment for details to corps: 3/6th NY (with the hope that the permanently assigned escort coys will mean that that's all you need)

And the rest is two triangular brigades of full-strength or nearly-full-units in a cavalry division, plus one regiment for south of the White Oak.

If McClellan could get the 1st NJ and 5th PA, then he has nine full size regiments (though can't form three triangular brigades because of the need to garrison south of the White Oak) but what I'm wondering is if you can also extract some more cavalry by raiding the forces under Dix around Fort Monroe. It looks like the 1st NY Mounted Rifles were basically one squadron in strength at this time (the other coys didn't organize until the July volunteer call), but for pure scouting that might be enough (or not) and if it is enough you could perhaps get the 11th PA off him. (Though it's moot with the arrival of the 4th PA, but it's too late by then.)

ED: but yes, if you want three brigades they either have to be binary, or possibly just made out of two regiments plus a tiny extra fragment.
Last edited:

67th Tigers

1st Lieutenant
Nov 10, 2006
Of course, the prime critic of not concentrating the entire force of cavalry from the army, and making it a completely independent body acting without coordination to the main army was Wesley Merritt. A typical example of his writings is his eulogy of Cooke.

Cooke was, of course, fired for incompetence, but Merritt refuses to accept this and asserts Cooke left of his own accord in disgust (only a month after the order firing him was published). In fact, what Merritt is trying to construct is a defence for Cooke - it was all McClellan's fault that Cooke turned in bad performances. In doing so, he pretends that the charge of the 5th US Cavalry at Gaines Mill was a glorious success that saved the army.

Merritt's writing is used by Wittenberg as an authority. Sadly, Merritt was a deeply partisan commentator, who objected to combined-arms action.


Feb 18, 2017
It looks like Merritt might actually be the source of the claim about cavalry being attached to "each corps, division, and even brigade commander"? Certainly he's a source for it, but gives no details that I can see.
It is admittedly interesting that Merritt completely elides Fair Oaks/Seven Pines.