If you want a specific answer to a specific line I don't answer here I'll provide it, but this is what I have to say on my chief doubts:
First, my apologies for not responding to each quote separately. I am a little strapped for time today, but I wanted to send this off while it is still fresh in my mind.
ELENNSAR:
And the evidence Thorp uses to support his position actually suggests less than that were present for duty, not his 75,000+.
Steiner said (copy/pasted from a text version of the document, so pardon any errors in my transcription)
"Wednesday, September 10. At four o clock this morning the rebel army began to move from our town, Jackson's force taking the advance. The movement continued until eight o clock P. M., occupying sixteen hours. The most liberal calculations could not give them more than 64,000 men."
and
"Thursday, September 11. General Hill s division, numbering about eight thousand men, marched through the streets, on their route westward, this morning."
https://archive.org/details/reportlewissteiner00steirich
That means that the highest number for those Steiner saw marching by - assuming his count is accurate, which is not something I would put money on under any circumstances given the difficulty of avoiding double counting - was 72,000. And quite possibly less. But on top of that, it is the total number of men marching by. It is not the number of men present for duty, but at best the army's aggregate present.
THORP:
Why do you think double counting occurred here? Didn't all of Lee's troops march through town just once on their way west when Steiner was counting them? Please give me some details as to how it is possible to double count troops that go by once? Wouldn't Steiner be just as likely to undercount in this circumstance?
ELENNSAR:
Per the return for September 22 (I could pick another for comparison's sake, but it is the closest to September 10 we have, and thus hopefully more representative of the number of men "present but not present for duty" than a figure from earlier or later): The army's "aggregate present" in those units whose strength is reported is 41,529 men. The present for duty figures for those units reported is 36,418. Subtracting Walker's division (which did not march through Frederick) for a more precise comparison, we have 32990 present for duty, 37649 aggregate present.
If the same percent of present for duty out of the aggregate present was true of Steiner's 72,000 tops figure, we'd have around 63,090 men present for duty out of those marching by, plus the troops not counted by Steiner.
THORP:
So basically, you are saying the that the Present for Duty to Aggregate Present ratio is 87.7% for Sept. 22, right?
ELENNSAR:
Stuart's cavalry (noted, at least in part, by Steiner, but whose numbers are not estimated) plus Walker's two brigades do not come anywhere near twelve thousand men present for duty.
THORP:
How do you arrive at the conclusion that the five brigades of Stuart's cavalry and Walker's brigade do not come "anywhere near twelve thousand men present for duty?" What are your sources to make such a statement?
Using John Allen's Sept. 22 Present for Duty numbers from Harsh's Sounding the Shadows:
5,159 (Walker) + 5,761 (Stuart) = 10,920. That is almost 11,000 present for duty.
Adding these troops not counted by Steiner to the total: 10,920 (Walker and Stuart) + 72,000 (Rest of Lee's army) = 82,920 aggregate present on Sept. 10 when Steiner recorded his observations, right? Is it also just another coincidence that the Baltimore Sun reported that in the timeframe between Sept. 4–6, "A Virginia captain, of much intelligence, told a relative in Leesburg during the recent passage of Lee's army through it, that the rebel force was there 84,000."
Using the Present for Duty to Aggregate Present ratio of your choosing, (87.7%) gives Lee 72,715 men Present for Duty. That is 23,700 men greater than the typical 50,000 men used by most historians, and larger than the number of troops McClellan had under his control for almost the first half of the campaign.
ELENNSAR:
And the other eye witness reports are all over the place, and some seem rather secondhand.
THORP:
Yes, all over the place, but the majority of which are higher. I have taken the time to source these eye witness reports. If you are going to dismiss them as "rather secondhand" then I would appreciate you specifically citing which accounts don't meet your muster and why.
ELENNSAR:
They cannot be used to confirm that Lee had not suffered an appreciable amount of men falling out of the ranks between September 2 and September 10, let alone between September 2 and September 17.
THORP:
Why not? What numbers and primary sources do you propose we use? What proof do you have that men fell out instead of catching up? The 2nd Manassas Campaign had disorganized Lee's army somewhat—like any campaign does—and Allen's numbers are taken from the end of it. From Sept. 2 to Sept. 10, Lee's numbers likely grew as he paused both at Leesburg and Frederick to regroup, resupply, recruit, and rest. Isn't it pretty well excepted that Lee gained 900 Maryland recruits during that time?
ELENNSAR:
Point #2: Thorp gives the figure of present for duty in the entire army (except for the reserve artillery and cavalry) as the infantry present for duty on September 22. As the September 22 return specifically includes to two battalions of artillery (Walton's and Lee's), and it is reasonable to assume that the divisions include the artillery attached to them as well as the infantry, one wonders how closely he actually paid attention to the return in question.
THORP:
The Sept. 22 return includes all of Lee's troops except for the Artillery Reserve and Cavalry, as already noted. You must observe though that the return is an undercount, or as Lee terms it, "very imperfect." 67th Tigers chart from yesterday does a great job showing how much of an anomaly the Sept. 22 report is.
Go to the the Oct. 10 return, the first to include both the Artillery Reserve and the Cavalry. It shows: Reserve Artillery: 912 and Cavalry Division: 5761, for a total of 6,673 Present for Duty.
To calculate apples to apples, Lee's army present for duty, exclusive of the Artillery Reserve and Cavalry, was 57,600 on Oct. 10, and 36,418 on Sept. 22, or 63.2% smaller, right?
So to get a fair and conservative estimate of the Artillery Reserve and Cavalry on Sept. 22, we take 63.2% of 6,673 (Artillery Reserve and Cavalry), which is 4,219, and add that back into 36,418 (Lee's army exclusive of Artillery Reserve and Cavalry) and we get a total Present for Duty on Sept. 22 of 40,637 Present for duty, or rounded up to the nearest thousandth, about 41,000 men. Does that satisfy your concern with how much I pay attention to the returns?
ELENNSAR:
Was Lee's army on September 2 larger Sears described it? Very probably (with limited records I'm reluctant to say "certainly"). But does that mean that McClellan actually had to deal with 70,000+ Rebels in Maryland? No. Thorp has given nothing to give that claim any more credibility than the opposite extreme of Lee having closer to 35,000 than 40,000 all totaled on September 17, and that claim can be poked full of holes without even relying on the latest research (even if possibly a dry read, it would be nice to see Harsh's work at least referenced as a more recent study than Sears in something written in 2012).
THORP:
I am not following how you arrived at your conclusion that McClellan did not have to deal with 70,000+ Confederates in Maryland. Did you not just write above that, "we'd have around 63,090 men present for duty out of those marching by, plus the troops not counted by Steiner." By your own calculations, after adding in Walker and Stuart, doesn't your number come to over 70.000? It would appear that you agreeing with my numbers then, right?
I don't understand your last sentance. Who should be referencing Harsh?
Bringing the