Richmond Depot Clothing – Volume II

Characteristics and Anomalies:
More Jackets, Pants, Drawers, and Shirts

By Richard M. Milstead
©2021 All Rights Reserved

Part 1: More Jackets
”Catching Up” Results from New Inspections of Original Examples

Jump to:

• John Blair Royall Jacket

• Gettysburg NPS 1st Sgt. Jacket

• Charles S. Tinges Jacket

• Second look at jackets in 2018 study

  • John Blair Royal jacket

  • Gettysburg NPS 1st Sgt. jacket

  • Charles Tinges jacket

• Four jackets not in the study group, newly inspected and photographed

  • Thomas D. Vredenburg Jacket

  • Capt. E. S. Marsh jacket

  • Charles A. Milhouse jacket

  • Allen C. Redwood jacket

The first three jackets which will be discussed were included in the comparison of fifteen examples that formed the basis for the Volume I analysis. The photographs presented in the earlier work were obtained from other researchers. I have now had the opportunity to personally inspect and photograph each of these jackets. Included here are “2nd – looks” at the John Blair Royal jacket, the Gettysburg NPS jacket, and the Charles Tinges jacket.

Four additional examples believed to be products of the RCB are analyzed that were not explicitly part of the comparison study group of fifteen in Volume I due to insufficient photographic evidence. They were, however, identified in the set of “probable” RD jackets also listed there. These are the Thomas Vredenburg jacket, the E. S. March jacket, the Charles Milhouse jacket, and the Allen C. Redwood jacket. Recent personal inspections and photographic studies now permit more detailed analysis of each.

As noted in Volume I, the John Blair Royall jacket has been considered for some time to be archetypical of RCB jackets from the mid 1862 through mid-1863 timeframe. The provenance of being worn by Royall when he was wounded at Chancellorsville definitely establishes it’s manufacture and issue date before the second quarter of 1863, more likely sometime in the fall of ’62 or winter ‘62/63, This was a period of significant strain on the resources and capabilities of RCB production due to major shortages in wool for domestic fabric manufacture in late 1862, increased requirements to resupply Lee’s army following the hard campaigning of the summer of 1862, and repeal of the commutation system in October 1962. Of special interest is the somewhat unusual wool-on-cotton textile used in the jacket’s body. Described in the earlier study as satinet, close inspection has revealed that it is instead a “broken twill” two over one weave. The presence of “as issued” red piping on the epaulettes is also notable. Given the prominence of the Richmond Howitzers in Richmond and the jacket’s dating, this is probably the result of RCB “customization” for well-placed units or officers. In other aspects the jacket largely conforms to the standard characteristics of RD type II jackets.

The analysis of the Royall jacket presented in Volume 1 was done largely based upon an extensive set of detailed photographs provided by researcher Tyler Putman. The jacket was also discussed in Jensen’s 1989 study. Generally, the result of this “second look” confirmed the conclusions of the earlier studies. The images here show the jacket in views not available before. One oddity noted is the almost “radical” profile of the front or the jacket. Most RD jackets (in fact most period civilian men’s coats) have a gentle curvature along the front which provides for stylistic chest fit. In civilian garments there was often padding in the coat which accentuated the chest curvature to create the so called “pigeon breast” look desired in men’s high styled dress coats. Like all RD jackets, however, there is no additional padding present in the Royall jacket. As such, the reason for the abruptness of the curvature and the matching curvature of the facing might suggest “custom” fitting at the time of assembly which also could support the theory of some “special” attention given to the “Howitzers’” jackets at the RCB. This is speculation, however, since in all other ways the pattern, construction, and assembly details are consistent with “standard” depot Jackets.

A closeup shows the piping on the epaulettes and how it was constructed. Red woolen cloth was wrapped around heavy cord and sandwiched between the two layers of the epaulette when sewed together. This technique instead of applying the trim on the outside clearly indicates the work was done during assembly not afterward.

The fabric used in the Royall jacket was closely examined. In Volume I it was described as satinet, which Nineteenth Century sources say was a wool-on-cotton fabric variant, woven to resemble an all-wool fabric on the face side. In satinets, the pattern is such that each strand of wool passes over several cotton warps before going under a single or paired warps. This was described as four, five, or even six “shaft twills” depending upon the number of warp strands passed over on the “face” by the wool weft.

The drawings seen above illustrate the weave patterns for plain, 2-over-1 twill, and the “satin” weave pattern used in satinet. Close inspection of both the “face” side and the “back” side (seen in the hole made by the shell that wounded Royall) shows that the actual weave is 2-over-1 where the cotton warps are paired, and the wool weft (fill) is a relatively heavy gauge yarn. However, a twill pattern (diagonal lines) is not obviously present. The Royall jacket textile is called a “broken twill,” a variant 2-over-1 weave that does not produce such lines and would tend to disrupt visual evidence of the weaving pattern. The combination of the heavy weft and the “broken twill” closely replicates the appearance of all wool fabric on the “face” side also achieved with a satinet but without using the “satin” weave. In effect, this fabric is “faux” satinet which itself is “faux” broadcloth.

During the first half of the war, the RCB dealt with a both Virginia woolen textile mills and Richmond dry goods establishments to obtain the wool-on-cotton fabric used in their uniform production. Before mid-1862 they purchased significant quantities of the supply they used on a “lot” basis, that is purchasing whatever the vendors had on hand. After that time, they were moving toward standard contract arrangements with a specific group of mills but “lot” purchasing continued especially in times of short supply. Records for “lot” purchases from dry goods merchants indicated a wide range of fabric types and colors. Some terms used to describe these fabrics were:” Jeans”, “Kentucky jeans”, “Tennessee jeans”, “Country jeans”, “Penitentiary jeans”, “plains”, “Kentucky plains”, “Washington plains”, “satinet”, “cassimere”, “N.C. cassimere”, “ribbed cassimere”, “fancy cassimere”, “cassinett”, “fulled cloth”, “Army cloth”, and “kersey.” Likewise, the color of the “goods,” when described, included “slate”, “cadet”, “grey”, “blue”, “light blue”, “sky blue”, “black”, “brown”, “gray & brown mixed”, “brown & black mixed”, “drab”, and “sand drab.” Generally, the delivery invoices for fabric obtained under later contracts did not specify the fabric type but where they did, jeans and cassimere predominate. It is possible that this is an example of a “lot” purchase made during the shortage at the end of 1862.

Mid war contract invoices list “grey woolens”, “cadet jeans”, and “drab cassimere” among the entries for what was provided. It is unclear if “cadet” refers to a specific blue-gray hue generally associated with the term today or simply a general reference for gray dyed material. “Drab” implies cloth made with undyed woolen yarn – cream colored or sheep’s gray - on natural or brown cotton warp that yielded cloth with a dull light grayish brown to yellowish color. The wool yarn in the Royall jacket fabric was most likely either “dyed in the wool” before spinning or “yarn dyed” before weaving with logwood. Depending upon the mordants used and the processing the original color would have been anything from a medium gray to almost black.

Given the history of this example, it was received by a 1st Sergeant in the ANV during the first half of 1863 if not in late 1862. In the analysis done for Volume I, photographs provided by researcher Dan Wambaugh were extensive enough to visually characterize its features and verified in discussions with Mr. Wambaugh as I prepared text discussing the jacket. Physically seeing and photographing the jacket led to several surprises. After close inspection, I do not believe that the English cloth is plain weave as I concluded earlier but rather it appears to be a 2-over-1 weave which, like the Royall jacket, may be a “broken twill.” The woolen fabric is lighter in weight that the typical kersey found in RBC products dating from the 1864-1865-time frame, similar to that in other 1863 examples such as the Bryan and Greer jackets. Furthermore, the plain-woven cotton lining is also lighter in weight than the osnaburg typically seen in RD jackets, more like a sheeting or shirting. RCB Inventories do show these materials were available and could have been substituted for linings. Visually, it still looks close to osnaburg.

The non-commissioned officer insignia is the infantry service color, blue. The chevrons were made from strips cut from heavy woolen cloth which appears to be of kersey weave and perhaps is of English origin. They have been neatly attached to sleeves perhaps by the sergeant himself, but also possibly by a “company tailor” in camp or a ”seamstress” to his specifications.

The images shown show closeups of various interior details of this jacket. Generally, this example conforms closely to characteristics described in Volume I. The facing shape is as is typically seen, and they are attached to the lining with a straight seam where the seam allowances are pressed toward the back. A slash pocket was constructed on the left breast lining. The body and lining (including facings) were sewn together from the neck opening all the way around, turned through the neck opening and top stitched (running stitch) around the periphery. The inside collar edge was left raw and overcast down to close the lining. The sleeves linings were also closed in the usual way. It is worth noting that the thread used in interior construction of the jacket was “whited brown” linen or undyed cotton.

Noncommissioned officer chevrons are rare on surviving Confederate jackets and this is the only RD example with original war period examples. In this slide the images show some closeups of one of the chevrons to illustrate the material they were made from and how they were constructed. The “Kersey twill” weave of the fabric strips is notable as is the heavy, uneven surface. The color, sky blue is consistent with the Infantry branch of service trim color. Such textile was imported through the blockade from England for making C.S. trousers throughout the war and is also similar to fabric used for Federal trousers and greatcoats.

The strips are relatively uniform in width but are, clearly, not bound on the edges. They have been attached with a running stitch catching the sleeve that may be undyed or, more probably, very faded vegetable or logwood dyed cotton showing a very faint brown hue. The chevron stripes extend almost to the seams on each side of the sleeve. The center lozenge is cut from whole cloth and not pieced.

As was discussed above, close inspection of the EAC in this jacket, shows that the weave is a 2-over-1 pattern, but the characteristic twill lines are not present. Based upon this, I believe that, like the fabric in the Royall jacket, this may be “broken twill.” It is less noticeable in this case because both warp and weft are of the same yarn. The difference between thickness (gauge) and color of the Royall fabric’s woolen “fill” and the cotton warp accentuated the effect on the appearance of the finished fabric. Closeups of weave in two different areas of the jacket illustrate this. Also, one area shown retains some degree the original napped finish while the other has been abraded exposing the weave pattern more distinctly.

The Tinges jacket is an excellent example of the RD type III jacket. Tinges served in the 4th Maryland Light Artillery (Chesapeake Artillery). His receipt of this jacket is documented in papers that came from the family when the jacket it was acquired by its current owner. Specifically, included with these papers was the duplicate (soldier’s) copy of his clothing roll for 1864 and 1865 kept by the battery’s commanding captain. This documents clothing issues to him in that period. He received a jacket on 9 December 1864 and another on 25 January 1865. The jacket’s condition supports late war production with little “in service” wear, strongly suggesting it was probably the last one he received. It is assumed that the period correct Maryland buttons on the jacket are war time substitutions (a common practice) but could be post war modifications as well.

The photographs used in the earlier analysis of the Tinges jacket were obtained from Les Jensen and Allen Wandling, the current owner. Two construction variations in the jacket were noted from recent personal study. It was observed in Volume I that the characteristic stitching line on each sleeve at the cuff was significantly higher up the sleeve than usual. This was confirmed from personal inspection and resulted from an unusual variation in the way the sleeve lining was secured to the outer sleeve. Also, while the interior was generally consistent with other RD III examples, the lining was attached to the body shell in a slightly different way. In other aspects the Tinges jacket exhibits typical RD III characteristics.

Shown above is Charles Tinges’ copy of clothing roll indicating what was issued to him for 1864 and 1865. These were kept by the battery (or company) commander for each soldier in their command as an accounting of clothing items issued by the army along with the cost for each item supplied. Confederate Army Regulations stated that “Such articles of clothing as the soldier may need will be issued to him. When the issues equal in value his allowance for the year, further issues are extra issues, to be charged to him on the next muster-roll. The money value of the clothing, and of each article of it, will be ascertained annually, and announced in orders from the War Department.” The total money allotted to a soldier was based upon the cost of standard yearly issues of uniforms/clothing, shoes (or boots), caps (or hats), socks. etc. he was supposed to receive as specified in Confederate Army regulations. Based upon the actual items drawn, if the soldier received less than the value of his clothing allotment, the difference was added to his pay.

This record indicates that Tinges was issued three jackets in the period between 4 February 1864 and 25 January 1865. The standard issue was one jacket per year. In the same period, he received two pairs of pants, one pair of drawers, one pair of shoes and two blankets. The standard issue for these items each year would be two pairs of pants, two pairs of drawers, and four pairs of shoes. Also, a soldier was supposed to receive three shirts, four pairs of stockings (socks), and one cap yearly. According to this record Tinges received no shirts, socks, headwear or other clothing or uniform items from the army in the period, but the two blankets were equivalent to his entire four-year allotment.

The reason why Tinges did not receive full allowances of items other than jackets, pants, and blankets nor, in fact, received additional shoes or any shirts or socks at all is obviously unknown. Supply problems could be one answer as could the possibility that Tinges was a very careful “steward” of his issued clothing so he could receive the difference in his pay. He also may have been a wealthy individual who was able to purchase what he needed or had relatives in Maryland who sent him items to supplant government issue clothing. We will never know. However, he did receive all the basic uniform components (jacket and pants) to which he was entitled plus an extra jacket to boot. Family oral history suggests that in later years (after the war) he claimed to have been charged $ 14.00 for that extra jacket.

The construction and interior details of the Tinges jacket are mostly consistent with RD type III characteristics. The lack of wear and, particularly, absence of staining in the lining probably reflects it sustained little “in service” wear. As is frequently seen, the facings have been pieced but their shape and method of attachment to the lining is typical with the seam allowance pressed to the back. Like most “kersey” weave EAC jackets the inside collar edge is left raw and overcast to the facings and lining at the neck. The “slash” pocket is in its usual position in the left breast and constructed in the typical fashion. The jacket and lining are hand sewn entirely in logwood or vegetable dyed thread and the main seams remain “tight” also indicating little wear.

One small anomaly in construction is the way the lining is attached at the hem of the jacket. Usually, the body of the jacket and the body lining with the facings attached are sewn together (right sides together) around the periphery from the neck opening down one side in front, along the jacket hem, and up the other front. The lining is then “turned” through the neck opening and topstitched all the way around. In this jacket, the front of the jacket body is sewn to the lining facings in the usual way (right sides together), but both the hem and the neck opening are left open. After the jacket is “turned” right side out, at the bottom hem the back was turned up along the edge. The lining edge was similarly turned under and overcast down along the bottom. The jacket was then topstitched around the periphery as usual. Basically, there is probably no reason for this variation other than this was how that particular “seamstress” made her jackets.

When this jacket was analyzed for Volume I, a difference was noted between it and the other examples in the study. On each sleeve there was a visible line of stitching several inches higher than the cuff topstitching typically seen on the other jackets. Personal inspection revealed that the sleeve lining in this example was attached in an unusual way based upon common construction methods used in most others. The sleeve lining on this jacket was attached at the cuff in a way similar to only one other example in the original study, the George Greer jacket in the American Civil War Museum Collection in Richmond VA. As noted in Volume I, on the Greer example each sleeve cuff end was turned up over the sleeve lining’s raw edge and they were fastened together by the top stitching along the cuff, keeping both sleeve and lining together firmly.

The closeup pictures on this slide illustrate how this was handled on the Tinges jacket. Like the Greer example, the sleeve lining has also been turned over the raw end edge of the sleeve lining, but in this case the sleeve extends much further up inside and was overcast down to the lining along its raw edge. To hold the lining firmly in place in relation to the outer sleeve, a very well executed line of what appear to be “back” stitches was executed several inches away from the cuff. This is the line noted in pictures of the sleeve from the outside. Top stitching along the bottom of the sleeve seen on the Greer and most other RD jackets was, therefore, unnecessary.

This extended length the is sleeve turned up inside is very unusual, suggesting either that the sleeve pieces in the “kit” were cut far too long or that the original sleeves were modified after the jacket was made to shorten them for better fit. This is thought to probably be the “as manufactured” construction not a later contemporaneous post issue alteration for several reasons. All the stitching appears original, the thread is consistent with the remainder of the jacket, and measurements of the sleeve length are quite consistent with those of many other RD jackets the author has studied. Furthermore, while post issue modification is not out of the question, there is no visible evidence of previous cuff fold lines or earlier topstitching that would suggest the sleeves were longer at any point in the past and shortened.

Richmond Depot Jackets - New Studies

In rough chronological production order

Four other jackets have now also been studied which at that time Volume was published were considered “probable” products of the RCB, but for which insufficient photographic information was available to include them in the comparison group. Generally, these examples will be discussed here with more attention to significant details in their pattern or construction than the previous three. Three of these examples are strongly believed to be products of the Richmond Clothing Bureau. The fourth, has a somewhat more equivocal attribution.

The jackets are presented in roughly the chronological order in which they were made and issued. The first is the Thomas D. Vredenburg jacket.

This jacket was given to Springfield, Ill. native Thomas D. Vredenburg, a 1st lieutenant in Company I, 10th Illinois U.S. Cavalry, by a Confederate woman who took pity on him while he was in a Shreveport jail following capture in the timeframe of Federal operations during the Vicksburg campaign. During that period Vredenburg’s regiment was active with Union forces operating in Louisiana and Arkansas but the precise date of his capture is unknown. He wore the jacket throughout his 13-month stay at Camp Ford prison in Tyler, Texas.

This example is made of wool-on-cotton jean cloth (brown cotton wrap; originally logwood dyed wool fill) with cotton osnaburg lining. It is an eight-button example and retains six of its original ten Gibson wooden buttons. It was machine top stitched using chain stitching, but the long (internal) seams were all hand stitched. The inside pocket is in the usual location but is unusual in that the opening is welted and overlaps the inside facing. Another unusual feature is the two-piece epaulette construction with cotton osnaburg underneath instead of the woolen body fabric.

The Vredenburg is a significant example of an early RD type II jacket for several reasons. It is the only example that has come to light made from wool-on-cotton fabric that retains Gibson wooden buttons still attached with original, undisturbed thread. Combined with its oral history, this says the jacket was made and issued in a relatively small window of time. Gibson button deliveries to the RCB did not start until the middle of November 1862. While the precise date Vredenburg received it is unknown, the latest date that could have occurred is sometime in June 1863. The Confederate history of the jacket is obviously also unknown, but the presence of Gibson buttons strongly suggests RCB manufacture. It could have been brought or sent home by an Arkansas, Louisiana, or Mississippi soldier from the ANV. This would, in turn, imply an issue date earlier, perhaps in late in 1862 but certainly sometime in the 1st Quarter of 1863. Either way it is one of the earliest known RD type II jackets and the earliest documented case of Gibson buttons used on any Richmond jacket.

Beyond dating the jacket, Gibson buttons provide important verification of RCB manufacture for a jacket which includes construction anomalies not seen in other examples. These will be discussed in more detail below but, taken as a whole, they clearly demonstrate the degree of variability present in the products of that operation. Gibson buttons, therefore, help in identification of this example despite the variability in construction details and provenance that would otherwise make such attribution, at least, equivocal.

The above images show an interesting inside construction detail of the Vredenburg jacket. The inside facings are one piece and the usual shape for a RD type II jacket, but they are attached to the body lining on each side by overlapping it and sewing them down along the raw edges using a ‘blanket stitch.’ More typically the facings on wool-on-cotton jackets are seamed to the lining directly (seam allowances pressed toward the back) or, if overlapped, the edge of the facing is turned under before being overcast to prevent fraying along the edge. While overcasting the raw edge on the facings is seen in jackets made of all wool EAC, it is unusual in a wool-on-cotton fabric jacket because of the tendency (as can be seen in the closeups) for such a loosely woven textile to unravel (fray) on the cut edge. Obviously, for this reason the maker chose the ‘blanket stitch’ technique which mitigates the effect of such fraying.

The inner collar was also secured in this fashion. The closeup from the inner neck closure shows the inner collar overlapping the lining along the raw edge to close the neck hole, also ‘blanket stitched’ in place.

Otherwise, the typical osnaburg lining is attached to the body shell as normal, sewn together around the periphery, turned through the neck opening, and topstitched all the way around. Similarly, the sleeve linings are set into the lining in the usual way.

The jacket is seen here from the back. As mentioned above, the construction of the epaulettes is unique among the RD type II examples reviewed in both terms of the textile composition and the construction technique used. As is shown, the face side is the made from the same wool-on-cotton jeans used in the body shell and sleeves, but the underside is cotton osnaburg. The edges on both the face side (jeans) and backing (osnaburg) were tuned under and the backing overcast in place with no topstitching. This construction method is also unusual in a jacket but is used for attaching the cotton “backing” on the adjustment belt for some RD pants (see Part 2). It is not known why osnaburg was substituted for underside woolen piece. While it is doubtful that the RCB “cutters” saved much jean cloth by replacing it with cotton, this could be a possibility. Another could be as simple as the “kit” this assembly worker received contained only two epaulette pieces instead of four, necessitating an improvised solution.

As no stitching is visible from the inside lining, the belt loops are attached by hand stitching to the body shell only, done before the lining was set in. This is seen on one other example, the Brunson jacket, described in Volume I. From the closeup, the hand stitching employed on long internal seams throughout the jacket is clearly visible.

The images on this slide show details of the inside pocket on the Vredenburg jacket. It is a welted pocket rather than the typical “slash” inside pocket found on most RD jackets and like those used on vests and the exterior jacket pockets on some Southern Depot jackets. The welt is made from the same osnaburg as the lining. As can be seen, it was also set into the left side of the lining such that it overlaps the edge of the left inside facing. This says that the left facing was attached to the front lining piece before the pocket was installed.

This pocket construction is unique in RD jackets with which the author is familiar both in being “welted” and in its placement overlapping the left facing. Functionally, however, it meets the apparent RCB requirement that pieceworkers assembling jackets provide an interior pocket in them. It is perhaps the most extreme example of a “seamstress choice” variation encountered in the entire study and, like the Ramsey jacket pocket described in Volume I, does show the “functional” rather than “technical” inspection acceptance philosophy at the RCB. That is, it is not like the normally made ones, but it does the job!

The use of sewing machines in the construction of RCB products was documented on two jackets in Volume I, the Bryan and Brunson examples and is reported by another researcher on the Abraham Adler jacket in the “probable” group, so the machine stitching is well documented. What is unusual here is that the machine stitching on the Vredenburg jacket, as shown in this slide, was done using a “chain” stitching machine.

Chain stitch sewing machines first appeared in the early 1830’s and pre-dated lock stitch type machines by at least a decade. Their principal disadvantage of chain stitch technology was that it could easily be unraveled causing the stitching to come loose. It was used in some applications, such as closing grain bags, where that tendency was useful to allow the seam to be quickly opened. However, for construction of clothing this was a significant disadvantage since seams holding the garment together could come apart easily. All long internal seams on this jacket were securely hand sewn. Using chain stitching for the topstitching was less of a problem because of this. Even so, several areas of the topstitching on this jacket have unraveled at some point in the past.

The wool-on-cotton type fabric used in the Vredenburg jacket was commonly used in large numbers of RD jackets and pants, particularly before the early part of 1864 when the decline of domestic textile mill output caused English woolen goods to dominate uniform production at the RCB. “Jeans cloth” was commonly called out in delivery invoices from local mills throughout the war. Early RCB documents refer to various types of jean cloth like “Kentucky jeans” and other variants which probably were “marketing” names, not different fabric types. One period source stated “[t]he local woolen mills of Kentucky… long enjoyed a reputation for the quality and quantity of the jeans which they produced, hence, the name Kentucky jeans has come to be generally applied to various grades of the cloth, whether made in Kentucky or elsewhere.” Jean cloth was made as a 2 over 1 twilled fabric. The specific weaving pattern for it is shown in this slide.

Colored fabrics were often either “yarn dyed”, that is the woolen yarn used in the fill was dyed after spinning into yarn, or “dyed in the wool,” where the woolen yarn was spun from a mix of pre dyed fibers to achieve the desired tint or shade. In such treatments, the wool was pre colored and the cotton “warp” color would be distinct from the “fill” color. The jean cloth used in the Vredenburg jacket was originally woven from brown dyed cotton warp and wool fill, most probably “dyed in the wool” to a gray shade using logwood. Because logwood is not light fast, uniforms made from these fabrics faded to a range of brown or even green in field service depending upon the mordants (chemical treatments before the wool is dyed) used and processing times. Some areas of the garment show fading of the wool weft to a light brown hue. In Volume I, the warp color of this jacket was incorrectly identified as natural (undyed). Inspection of the garment itself revised this conclusion. In most areas, the warp yarn is a mellow brown. Only where significant surface abrasion has occurred to the garment, is the natural color of the cotton yarn exposed. The photographs and closeups here show these effects.

The original fabric’s surface was also originally lightly “napped” as seen in a less exposed area where the surface abrasion is not as severe.

Captain Marsh was wearing the jacket when wounded at Chancellorsville in May 1863. Because of the severity of his wounds, he did not return to active service with the regiment. The jacket, therefore, probably dates from the 1st half of 1863 or possibly even the last quarter of 1862. Like several other examples studied, this is an early case of officer use of an enlisted jacket for field wear. This jacket is made from “kersey” weave English Army Cloth with 9 buttons on the front. Those buttons and the ones securing the epaulettes are imported script “I” buttons marked “S Isaacs, Campbell & Co.” They are possibly the buttons originally provided on the coat as manufactured, however, most have been re-attached using different thread. Given the probable period of the jacket’s manufacture, these could be from large consignments of military supplies purchased starting as early as the summer of 1861 by Confederate Ordinance Captain Caleb Huse in England from S. Issacs, Campbell & Company and run through the blockade. Ship manifests in late 1861 and in 1862 list large quantities of “brass buttons for infantry, artillery, and cavalry” or simply “army buttons.” The kersey EAC from which the jacket was constructed likely was also purchased through this company as such cloth is also listed in cargos received through the Blockade in this period. The Haines jacket discussed in Volume I is another example of kersey woven EAC used in the 1862 timeframe. The lining is the typical osnaburg used in RD type II examples.

The epaulettes on the example are one piece with the edges turned under but have the same rounded end shape usually seen in the two-piece epaulettes found on other early RD type II examples. Like the Vredenburg jacket, this may be a case of a “seamstress choice” decision to work around a deficit in the pieces she received in the “kit” for this jacket. These are different in both pattern and implementation from single piece epaulettes discussed in Volume I on the Ramsey and other late RD type II jackets (for example, the Milhouse jacket discussed next) which are shaped differently, having a definite point at the end rather than being rounded.

Marsh, or someone he hired, crudely added three strips of gold braid to each side of the collar to signify his rank but seems to have made no further modifications to an otherwise standard issue private’s jacket.

In the image above, the Marsh jacket is seen from the front. Closeups show its buttons’ front and back, including the S. Issacs, Campbell, & Co. marking. They are currently attached with a non-period gray thread also used in other conservation work done to the jacket implying they were sewn back at the same time. All eleven are identical to war period examples imported through the Union blockade strongly suggesting that they are original to the jacket despite the reattachment. Another closeup shows the gold braid added to the jacket’s collar denoting Captain’s rank.

The other closeup images focus on the jacket’s epaulettes. As can be seen they are constructed from single layer of the body shell kersey EAC where the edges (seam allowances) are turned under and secured with a running stitch which appears as topstitching from the top. They are also “functional” in that they have buttonholes allowing the epaulettes to be unbuttoned. The thread used in binding them appears to be period suggesting this could be the work of the “seamstress” who made the jacket or, at least an enhancement Marsh had done, perhaps, when the rank insignia was added. In any event, this is the only example of “functional” epaulettes seen in any of the RD type II jackets studied to this point.

The back of the Marsh jacket is shown in the above image. The pattern, construction and implementation details are typical of many of the RD type II examples. The jacket is entirely hand sewn using (now faded) logwood dyed thread. The topstitching was done with running stitches also typical of many other examples. Perhaps the only anomaly is the “whimsical” treatment of the belt loops with pointed ends, which are totally unique to this jacket.

The lining and facings, pocket, and construction details inside this jacket are also characteristic of most RD type II examples (see Volume I) as can be seen in the closeups. The facings are the normal shape and attached to the lining with the seam allowances pressed to the back. Then lining/facing assembly is attached to the body shell around the periphery from the neck opening with a line of running stitches. It is then turned through the neck hole and topstitched (again running stitches) all the way around. The pocket as usual is a slash type located on the left breast.

This jacket was worn by Charles A. Milhouse, a member of Co K, First Confederate Engineer Regiment. Milhouse joined in South Carolina in January 1864 and apparently served in the Department of Virginia for the remainder of the war. It is now in the collection of the South Carolina Confederate Relic Room and Military Museum in Columbia SC. Milhouse may have received it when released from hospital following an illness in August 1864. Besides obviously hard use in the service, the jacket retains none of its original buttons and also has suffered significant insect damage.

This example exhibits some of the RD Type II/RD type III “transition” characteristics noted in the Ramsey jacket (discussed in detail in Volume I) such as an “angular” collar and “pointed” single piece epaulettes. It has an 8-button front and is made of plain (“tabby”) weave EAC. Such fabric is also seen in the George Wilson pants in the Smithsonian Collection (discussed in Part 2) which dates from the second half of 1864. The hand sewing on this jacket appears to be somewhat more crudely executed than usually encountered. Also, the belt loops were made from different fabric than the rest of the jacket.

Unfortunately, the Milhouse jacket is in very fragile condition due to extensive insect damage on top of field “wear and tear.” In this slide, the angular collar and epaulette shapes are shown. It is believed RD type II jackets made until roughly the end of the first quarter of 1864 exhibited collars which were rounded at the front neck opening. Before the type III jackets began to be produced, a subtle change appears to have been made in the type II collar design to make it “angular” rather than rounded. Several known or suspected RCB made examples, probably issued in the June through August period, exhibit this. The design is termed “transition” because such “angular” collars also appear on many surviving RD type IIIs that can be documented to have been received in the field in fall 1864 and winter ‘64/’65. Given lag between manufacture and issue in the field, it is speculated that these “transition” type II jackets probably began appearing in the latter part of the second quarter of 1864. Also, it is doubtful that there was a sharp demarcation point in field issues. That is, very possibly both the “transition” RD type II and RD type III jackets were being issued to troops at the same time, perhaps even were both distributed at the same time within the same unit.

Another feature of these late “transition” RD type II jackets seems to have been a change in both the shape and construction of the epaulettes. It cannot be said with certainty that both changes were simultaneous but all three RD type II jackets displaying the “angular” collar have “one-piece, pointed” epaulettes as well. The shape and construction of these are shown in the closeups above. Comparison with the epaulettes on the Blair Royall, Vredenburg, and Marsh jackets illustrates the difference. With the elimination of epaulettes on the RD type IIIs, this became a moot point.

The hole in the end of the epaulette seen in the closeup could either be insect damage or possibly where the button tore through at some point. As noted, the stitching throughout the jacket, but particularly the topstitching, appears rushed if not somewhat crude. The pieceworkers who sewed together uniform clothing for the RCB were usually quite competent and the stitching more precise.

Interior details of the jacket are shown in closeup images above. The interior construction is generally typical. The facings are attached directly with “running” stitching and the seam allowances pressed back. The inner collar overlaps the lining at the neck opening. As seen in the picture of the neck area, its edge was left raw and overcast down without being turned under, a common method in jackets made from EAC.

The topstitching shown from both the front and back sides at the left front bottom (where the curving occurs) again displays somewhat crudely executed “running” stitches used throughout the jacket. This is also seen in the picture of the jacket cuff. The fraying visible along the edge of the cuff, demonstrates the significant “wear and tear” the jacket has sustained in service. This fraying is also seen in the closeup looking inside the cuff. The sleeve lining was detached at some point and has been sewn back using non period thread, possibly the during conservation work to preserve the jacket after the war. Further damage, most likely from Milhouse’s use, is evidenced in the interior pocket which has been ripped and the pocket bag detached.

The Milhouse jacket is seen here from the back. The construction is typical. It still has belt loops despite being a late RD type II with an “angular” collar and one piece, “pointed” epaulettes. Interestingly the belt loops are constructed from EAC which does not match the rest of the jacket. The hew is different and unlike that in the remainder of the jacket body, it is “kersey” weave EAC displaying visible diagonal twill lines.

The Milhouse jacket is made from plain (aka “tabby”) weave English cloth. This image is from the jacket sleeve where a small break in the stitching has occurred. Closeups show the plain weave fabric and the diagram illustrates its weaving pattern. The white color of the clean, unstained cotton osnaburg used in the sleeve lining (also plain weave) is seen in another closeup.

Plain woven EAC is unusual but not unheard of. The same (or remarkably similar) textile is used in the George Wilson pants in the National Museum of American History collection, Smithsonian Institution in Washington DC. These RD pants will be discussed at greater length in Part 2 of this presentation.

Allen C. Redwood originally served with Company C of the 55th Virginia Infantry, raising temporarily to the rank of Sergeant-Major in 1863. He transferred to the 1st MD Cavalry as a private in February 1864 serving with them until captured in April 1865. No records exist of clothing issues to him during that period. Like the jacket belonging to his brother, Henry, this example is pictured in the Time/Life series volume Echoes of Glory - Arms and Equipment of the Confederacy. Redwood later became a famous artist who did many illustrations for Battles and Leaders of the Civil War and other publications.

The Redwood jacket is made from the typical blue-gray kersey weave EAC seen in the Tinges jacket discussed earlier and other RD type III jackets studied in Volume I. It also has a cotton osnaburg lining and appears entirely hand sewn. The jacket has an eight-button front like the Knight jacket, one of the RD type IIIs discussed in Volume I. Also like the Knight and Tinges jackets, its buttons are Maryland state buttons contemporaneous to the war period. Several RD type II jackets have eight button fronts including the Vredenburg and Millhouse examples described earlier, and the Abraham Adler jacket presented in Volume I as probable RCB made jacket.

Details in the way this jacket was assembled are atypical in RD jackets, however, and are closer to the way Federal jackets were constructed. The construction of the collar is unique among RD jackets studied by the author. The usual topstitching around the hem, on the edge of the collar, and around the cuffs is absent but is seen along the front on both sides starting at the collar and terminating abruptly at the hem. The front on both sides is squared off at the bottom not curved as is unusual. Only one other RD type III, the John Henry jacket exhibits this characteristic. It has a seven-button front. In his 1989 MC&H article Jensen states that it was altered at some point and may have been shortened.

The proposition that the Allen Redwood jacket was altered, perhaps to shorten it (like the Henry example), is supported by the construction at the bottom front corners. However, when combined with the other anomalies, such deviations from typical RD jacket assembly methods could instead suggest that the jacket wasn’t produced through the RCB at all but is, instead, a privately made jacket. Somewhat less likely, the jacket could have been both privately made (perhaps for someone else?) and Redwood subsequently had it further modified to adjust its length for his own use.

As discussed above, these images show several details of the Allen Redwood jacket (seen from the front) that reflect construction variations from other RD type III examples. The lack of topstitching around the periphery of the collar and at the cuffs is unusual. The jacket is “squared off” at the intersection of the front and jacket hem as opposed to the rounded corner typical of most RD jackets. The replacement Maryland state buttons, while not rare on jackets with a Maryland soldier provenance, are also noted as likely modifications made at some point in its history.

The above image shows the Redwood jacket viewed from the back. The back panel pattern is quite typical of RD type III jackets discussed in both Volumes, which, like this one, are usually, straight hem across the back in the center. The view looking inside of the cuff shows fraying along the edge, a common wear point noticed in many jackets, but the lack of topstitching at the end allows this to gap significantly, perhaps suggesting the reason for doing so during assembly on others.

The topstitching typically found around the hem is also not present. As discussed below, this method of attachment of the lining to the body shell hem is not unique but is, nevertheless, unusual among the RD jackets studied.

The image of the inside of the jacket illustrates other noteworthy details in its construction. The “slash” pocket is handled in a manner like most other jackets and as is attachment of the sleeve linings. However, other construction details are very unusual.

First, the lack of topstitching around the collar noted above is explained by the method of construction seen in the closeup of the right-side inner collar. The usual method of assembling the collar pieces (outer and inner) in RD jackets is by sewing them together (right side to right side), turning them and pressing the seam down. Topstitching added around the edge secures the seam holding the two collar pieces together once the outer collar is attached to the jacket body shell. The inner collar is turned down and the neck closed along the lining using one of three different methods discussed in Part 3 of Volume 1. On the Allen Redwood jacket the outer collar was first attached to the neck opening of the body shell and its top and end edges pressed down. After the lining was set, the inner collar was overcast to the outside edge of the outer collar as well as over the lining at the neck opening. The edges of the inner collar were left raw all around. This method of assembly is used in construction of many Federal uniform jackets and coats, particularly those made at the Schuylkill Arsenal, although some contract made jackets’ inner collars are done in the same fashion, only machine stitched rather than overcast into place. It is also a technique used in civilian garments. Generally. Federal examples, like this jacket, do not have topstitching around the edge of the collar either. No other RD jackets reviewed in these studies have been assembled by this technique.

Second, it is noted from the closeup of the left facing, that the facings were sewn to the front in the way conventionally seen on other RD jackets, that is, sewn right sides together, turned, pressed, and topstitched. However, the seam joining the two extends only down the front on each side, as does the topstitching which is abruptly ended at the corner. The next slide will show how the bottom of the facings are handled.

Third, the lining was overcast to the body shell at the hem in a similar manner to the Tinges jacket, but without topstitching done on that example along the hem to further secure it to the body shell. This is also like what is seen in Federal jackets except usually they have a bottom facing overcast to the body shell to which the lining is attached.

Finally, in a detail which may be significant, thread used in sewing different areas on the interior appears to have faded differently since original application. That is, while the weight and fiber composition of all the thread used appears similar, thread used in the buttonholes and topstitching down the front appears to have faded more completely that that used in overcast stitching at the collar, along the facings and at the hem. Logwood dyed thread used during the period faded through exposure to light and heat, but the variation in degree observed here could possibly indicate that the stitching was done at different times.

The construction at the bottom corners in the front of the Allen Redwood jacket is shown in this set of images. This method also is unique among the jackets studied. The bottom edge of the facing and hem of the body shell have been turned under (up) and the facing folded over. On the facing’s back the raw edge was overcast down from neck to hem over the lining which runs under it. There is also hidden stitching securing the facing to the body shell at the bottom along the hem perhaps done before the facing was attached to the lining. From areas where there is insect damage on each facing, it appears that the lining extends all the way under these facings as an interfacing for the front of the jacket.

While such construction could simply be indicative of just another variation (anomaly) in the assembly by this specific pieceworker, it could have another explanation. Shortening of jackets to obtain a better (more fashionable) fit is one of the post issue modifications made by soldiers, or others they employed. As noted above, another RD type III example, the John Henry jacket, apparently was shortened by a seamstress. Unfortunately, the author has not had the opportunity to inspect that example to see how the modification was made in that case. Simply based only upon the way the ends of the facings have been treated, however, no absolute conclusion can be reached on whether this jacket was shortened or made this way originally.

Some further perspective on the Redwood jacket is provided through comparison of dimensions, taken the day it was examined, to other relevant RD examples studied over the last few years. The table above shows three specific measurements from all the RD type IIIs and the two 8 button RD type IIs that have been examined. The three measurements represent the distance on each side from the neck opening to the hem in front to where the corner curve ends (i.e., neck to hem in front) and the distance from the base of the collar (where the collar center seam meets the neck seam) to the hem along the back center seam. Unfortunately, in these studies, measurement precision was subject to external factors when each jacket was examined, not the least being the amount of time available to collect these data and/or the ease with which I could manipulate the jacket to obtain accurate measurements. In general, even with small differences, the relative values provide useful information. What is true here is that among RD type III examples, the Redwood jacket appears to be clearly the shortest in body length on the back seam from collar to hem.* Combined with how the inside facings were handled (discussed above), the dimensions of this jacket strongly suggest that it was, in fact, shortened at some point.

But was the collar modified at the same time? There are many examples of customization of RD jackets running from changing buttons, to cutting off or removing epaulettes on RD type IIs, to adding additional (interior) pockets, or even to replacing linings, some done by hired seamstresses. In addition to John Henry’s jacket, Thomas Tolson’s jacket, also from the 2nd MD Infantry, (Volume 1) was heavily modified as were his pants which are described in Part 2. From those examples, some Maryland soldiers apparently did have personal customization done to their uniforms. Unfortunately, the Redwood jacket collar measurements were within the range of other RD type IIIs suggesting it wasn’t altered to lower it and doing so would have required disassembly of an existing RD jacket’s collar to remake it for no other apparent reason. Given this, the collar construction certainly could be original and merely another (very unusual) “seamstress choice” variation anomaly.

Alternately, could this jacket be a tailor-made RD jacket copy? This could explain the collar construction. Such jackets are known to have existed and there are pictures of members of both the 2nd MD Infantry and the 1st MD Cavalry (the unit that Redwood joined) wearing nonissue uniform jackets likely made by local tailors or seamstresses. Arguing against this speculation, however, is the facing attachment method down the front of the jacket with abruptly terminated topstitching and the color difference in the thread in that area with that used elsewhere inside the jacket. If it was originally a tailor-made garment, did Redwood, after its acquisition, decide to have it further modified to shorten it for better fit?

Unfortunately, none of these questions were answered based upon physical inspection of the garment, so the Redwood jacket’s “origin mystery” remains unsolved.

*Relative to the Knight jacket, closest analog to the Allen Redwood example, on the day I examined it, its owner would not allow me to unbutton it which prevented measurement of the right front and, possibly, affected accurate measurement of the left side as well. The back measurement on that jacket is also suspect. Many of the seams on this jacket had broken over time and it was kept together in some places using the expedient of safety pins! The neck seam at the back where the outer collar attached was one of those. In the rush of taking my measurements, under the scrutiny of the very, nervous owner, could I have gotten it wrong? I must concede this is probably not an accurate dimension either. The Vredenburg and Milhouse 8 button, RD type 2 jackets are very consistent dimensionally which further supports the probability that dimensions taken off the Knight jacket likely were incorrect.

End of Part 1