Here is the breakdown that Harrison and Busey give (from pages 457-461):Okay, I found one source: the official report of Major Charles S. Peyton of the 19th Virginia Infantry, who wrote the brigade after-battle report, as he was acting commander because of the high number of ranking officer casualties. His report states: "The brigade went into action with 1,287 men and about 140 officers, as shown by the report of the previous evening, and sustained a loss, as the list of casualties will show, of 941 killed, wounded, and missing, and it is feared, from all the information received, that the majority (those reported missing) are either killed or wounded."
I'll try to get more sources, as his report could be off.
Garnett and staff: 4 officers, 2 privates
8th Virginia: 31 officers, 49 noncommissioned officers, 162 privates
18th Virginia: 31 officers, 64 noncommissioned officers, 276 privates
19th Virginia: 38 officers, 76 noncommissioned officers, 312 privates
28th Virginia: 24 officers, 61 noncommissioned officers, 291 privates
56th Virginia: 28 officers, 67 noncommissioned officers, 335 privates
Total: 156 officers, 317 noncommissioned officers, 1378 privates
Here are the casualties they use:
Garnett and staff: 1 officer killed, 1 officer wounded, 1 officer wounded and captured, 1 officer wounded
8th Virginia: 2 officers/6 NCOs/26 privates killed, 7 officers/6 NCOs/20 privates wounded, 9 officers/8 NCOs/32 privates wounded and captured, 5 officers/16 NCOs/31 privates captured (total of 23 officers, 36 NCOs, and 109 privates as casualties)
18th Virginia: 3 officers/11 NCOs/30 privates killed, 6 officers/12 NCOs/37 privates wounded, 12 officers/22 NCOs/50 privates wounded and captured, 2 officers/3 NCOs/43 privates captured (total of 23 officers, 48 NCOs, and 160 privates as casualties)
19th Virginia: 3 officers/8 NCOs/29 privates killed, 9 officers/15 NCOs/20 privates wounded, 10 officers/12 NCOs/27 privates wounded and captured, 1 officer/10 NCOs/24 privates captured (total of 23 officers, 45 NCOs, and 100 privates as casualties)
28th Virginia: 2 officers/9 NCOs/24 privates killed, 4 officers/14 NCOs/27 privates wounded, 3 officers/11 NCOs/36 privates wounded and captured, 5 officers/6 NCOs/31 privates captured (total of 14 officers, 39 NCOs, and 118 privates as casualties)
56th Virginia: 4 officers/7 NCOs/16 privates killed, 2 officers/7 NCOs/27 privates wounded, 5 officers/8 NCOs/40 privates wounded and captured, 6 officers/9 NCOs/32 privates captured (total of 17 officers, 31 NCOs, 115 privates as casualties)
Total: 15 officers/41 NCOs/125 privates killed, 29 officers/53 NCOs/131 privates wounded, 40 officers/61 NCOs/185 privates wounded and captured, 20 officers/44 NCOs/161 privates captured (total of 104 officers, 199 NCOs, and 602 privates as casualties)
This was Harrison and Busey's methodology for coming up with these numbers and the roster that follows (from page 164-5):
"The individual unit rosters which follow were compiled from contemporary Confederate company muster rolls on file in the National Archives, in Washington, DC. The rolls used were primarily those of the April/June 1863 muster period which was conducted by the division on 30 June at and near Chambersburg, Pennsylvania. In a small number of instances, unit July/August 1863 muster rolls were used when 30 June rolls could not be located. In rare instances--such as with the 38th Virginia of Armistead's Brigade--when muster rolls were not available, either of the pre- or post-battle period, the author conducted a complete review of all individual compiled service records of soldiers who served with the unit during the entire course of the war. Every effort was made in reviewing these records to determine if each soldier was or could have been at the battle. Those who could definitely be determined to have been absent from the engagement were deleted and the residue were included in the roster. While this may at first seem to be a relatively inaccurate way of determining those present for such units, it is, given the current condition of Confederate records, virtually the only way a roster for the units could be compiled. All unit rosters were correlated against the personal recollections of participants in the assault which are contained in the narrative of this book. Only in rare instances did the roster fail to match the recollections of survivors. While it is virtually certain that some men listed in this roster were not at Gettysburg, and that some who were in the assault may not in fact be listed, the author is confident that it is as complete and accurate as is currently possible, given the passage of 129 years and current condition of Southern war records."
Later, she describes how any number of walking wounded could be unaccounted for as they either never visited a hospital or stayed for so short a time that it was not considered significant. Harrison speculates that as many as 300 wounded could be unaccounted for in this way (page 165).