Richmond Depot Clothing – Volume II

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

By Richard M. Milstead
©2021 All Rights Reserved

Part 3: Other Richmond Depot Soldier’s Clothing

Drawers and Shirts

Jump to:

• Introduction

• Enlisted Men’s Undergarments

• Richmond Depot Drawers – Comparison of Known Examples

  • Henry Hollyday Drawers

  • Lt. John Selden Drawers

  • Richmond Depot Drawers – Characteristics

The Quartermaster Department in Richmond manufactured or procured clothing other than just jackets and pants for soldiers. Drawers, shirts, socks, shoes, and uniform caps were predominant among these. Less is known about the first four than of headgear even though, in sheer numbers, those represented substantially more of the production activity and output. Part of the reason for this is because of the numbers of each that survived after the war. Like jackets, to veterans preserving their “army cap” or hat was more symbolic of their service than, for example, saving their drawers or socks. Drawers, shirts, socks, and shoes were expendable supplies needed to survive in the field but wore out quickly and the government issue versions were mundane enough not to be desirable keepsakes. Shirts, when preserved, often were special examples made by the womenfolk at home and carried personal significance for soldiers in their later years. The same was true of “homemade” socks. Examples of shoes more often survived as souvenirs preserved by Union soldiers than of the Southrons who wore them.

Part 3 of this presentation will focus on two of the least seen and studied items, government issue drawers and shirts made at the RCB and issued by AQMs in the field. The numbers produced were truly staggering. In April 1863, the RCB forecast that in the next year ending 30 April 1864 it would produce 275,000 shirts and a like number of pairs of drawers. In the period including last six months of 1864 through the end of January 1865, the QM Department issued to the ANV almost 158,000 cotton shirts, 21,000 flannel shirts, and more than 170,000 pairs of drawers, most of which were made at the RCB. Despite this, few original examples with provenance suggesting QM manufacture have survived or been identified as such.

In this study, a few that have will be described and analyzed to understand their pattern, material (fabric and notions), and construction characteristics. The author is aware of only two pairs of drawers and one shirt likely produced at the RCB which have come to light and can be so identified. One pair of drawers and the shirt are part of the ensemble of artifacts owned by Henry Hollyday which are in the Michael D. Kramer collection. Like Hollyday’s pants discussed in Part 2 of this presentation, they were part of a late war issue of clothing he received. A second pair of drawers owned by Lt. John Selden are themselves part of a clothing ensemble in the ACWM collection. His pants also were discussed in Part 2.

With such an exceedingly small sample size, general conclusions relating to the hundreds of thousands of each issued during the war are difficult. Simply the fact that they probably also represent only late war issue says they may not be entirely representative of early war products. However, by their very nature these are simple garments and as the RCB did not appear to make sweeping changes in either jackets or pants, so it is a reasonable assumption that they closely represent those made earlier as well.

The format will again be somewhat different from the first two parts of this presentation. The discussion of drawers will again be in the form of a comparison study, looking at specific areas of the two examples to note similarities and differences. With only two in the study group, the concept of “anomalies” used in earlier work really has less meaning. The discussion of the shirt will be even more focused, and the emphasis will be on the pattern and construction of that example and how it fits into the spectrum of other military and non-military examples of the period.

• Clothing worn under the uniform of the Civil War Soldier

  • 19th Century “undergarments”

  • History in the US Army

• The C.S. Regulations – What were soldiers supposed to receive?

  • Diversity & Quantity

  • Sources

Shirts and drawers are what were generally described as “undergarments” in most of the 19th Century. Not considered a fashionable element of men’s dress, they were not displayed in “polite company” except under other garments. Both were worn from before the 18th Century but evolved with the styles of men’s clothing.

The basic pattern of men’s shirts changed little for at least two centuries prior to the Civil War. In the 100 years before 1860, men’s everyday shirts tended to be simpler in form and, since most were made at home, changed less than dress styles over that time. For dress shirts embellishments such as ruffles and, in the first part of the 19th Century, pleated inserts, called “bosoms,” were added to enhance the chest and neck area as waistcoats and vests opened showing more of the man’s neck and chest area. Most common shirts had plain fronts and, until well into the 19th century, did not have separate plackets at the body opening. Until the 1850’s both dress and common shirts used the so called “square cut” pattern where the body was rectangular in shape made of one or two pieces of fabric, the sleeves either rectangular or trapezoidal, and inserted “gussets” provided expansion needed to provide for fit at the sleeves and neck. Styles and forms of collars and cuffs changed as did how extensively the neck and sleeves were gathered but the basic shirt form remained the same. Starting in the years immediately preceding the war, a new more fitted pattern, termed the “French style,” started appearing but most war era common shirts still followed the earlier pattern.

Drawers, or an undergarment worn beneath the outer clothing covering men’s legs, were used as early as the 17th Century. In the late 18th Century when breeches were in fashion, these were knee length, secured by drawstrings at the waist and knees and made from linen, cotton, or wool for winter. As the styles of men’s leg coverings changed to longer and more fitted, the shape of drawers followed.

Shirts and drawers were issued in the United States Army long before the Civil War. Shirts were made at the United States Clothing Establishment at Schuylkill Arsenal in Philadelphia early in the 19th Century. The Army started issuing drawers as well in the mid-1820’s. By the eve of the Civil War, both garments, made to standardized patterns, were being produced at Schuylkill Arsenal and issued to all enlisted men. The Federal Government issued them by the hundreds of thousands during the war. They were produced both at Government facilities and contracted from various private firms.

Following Secession, the Confederate Army’s regulations closely mimicked the U.S. regulations for clothing items issued to its soldiers. In a three-year enlistment an enlisted man was supposed to receive 9 shirts and 7 pairs of drawers. Throughout the war these were specified to be made from flannel. In practice, while wool or flannel shirts were issued sporadically, the vast majority were either cotton or unspecified (and likely cotton). In terms of frequency, it is almost impossible to correlate issue documentation to individual soldiers but from what is seen in the records, shirts and drawers along with socks and shoes represent most of the items issued to them. Moreover, the quantities issued in late 1863 through the beginning of 1865, based upon the author’s investigations, suggests strongly that if not plentiful, then, at least, they were available in reasonable quantities.

Issued shirts and drawers came from multiple sources. The Richmond QM Department received not only RCB produced supplies, but shipments of shirts were imported through the blockade. States like North Carolina and Georgia also made these for issue to their troops like uniform jackets and pants. Clearly home front Aid Societies and relatives at home made their contributions to the troops as well.

• Examples in the Study

• Henry Hollyday drawers

• Lt. John Selden drawers

Part 3 will begin with a discussion of drawers. In the following section, a comparison study will be presented of two pairs of drawers with provenance linking them not only to the Virginia theater of the war but to other items believed to be related to the RCB. They were owned by Henry Hollyday and Lt. John Selden also owners of pants discussed in Part 2. The study involved two separate personal inspections of the Hollyday drawers and one of the Selden ones and used 80 photographs taken by the author and eighteen more from other sources.

Critical to identification of the Hollyday example as a RCB product was the presence of three of four original Gibson contract “pant” size buttons. The significance of this cannot be overstated in my opinion. On a generic garment such as a pair of drawers or Hollyday’s shirt which will be discussed later in Part 3, little useful photographic evidence or other corroboration beyond provenance is available to tie them to a specific manufacturing source. Pattern similarity cannot be established until known examples are identified and, even then, for such simple garments, may be difficult to recognize. Fabric analysis is, of course, helpful but also inconclusive.

We are fortunate, therefore, that the products of John and George Gibson’s “House Carpentry and Building” establishment are present, a “Rosetta Stone” to decipher the work of the RCB.

The Henry Hollyday drawers shown in the above image were part of the same clothing issue as his pants (Part 2) and the shirt that will be discussed later. They are made from a common type of undyed cotton osnaburg which is similar to the lining material found in RD jackets and pants. As emphasized above they still retain 3 of 4 Gibson contract “pant” size wooden buttons. All are still attached with the undisturbed, original, undyed thread. Only the waistband button is missing. From examination the thread used appears to be cotton, but no fiber analysis was performed.

One of the remarkable aspects of this garment is its excellent, perhaps unused condition. There are no rips or holes on the material and all stitching (except for the missing button) remains in place. There is little staining discoloration except for some dark, splotchy marks on the backs of the legs that could be blood but look more like paint smears. Given what would be expected for undergarments from “hard” service in the field these are remarkably pristine.

We do not know how Selden procured these drawers which are part of the ensemble of garments and other artifacts in the ACWM collection that were owned by him. By regulation, he was supposed to purchase such items from the AQM serving his Artillery battalion or acquire them privately. He also could have simply been given them from available stock late in the war when restrictions against officers receiving government clothing were less stringent. How distribution of non-uniform items like undergarments to officers occurred probably varied. His shirt, also among the surviving pieces in the ACWM collection, is a British army shirt imported through the Blockade late in the war, probably intended for issue to enlisted men as well. In any event, no records have been found suggesting how he obtained them.

These drawers are made of the same cotton osnaburg fabric as the Hollyday example. Instead of Gibson wooden buttons these have white glass buttons often found on period civilian clothing. It is unclear whether these are replacements or original. The thread with which they are attached appears to be cotton and is consistent with that used elsewhere in the garment. There is also no obvious evidence that it has replaced other earlier stitching. Such civilian buttons do appear in various RCB inventories and purchase records throughout the war, so it can be assumed that they were used on government clothing. One factor mitigating this assumption, however, is that only two of the four original buttons are still present. At some point the buttons normally securing closure of the leg at the cuff were removed and replaced by tape ties instead. Selden or someone working for him modified the drawers to his personal preference, so the buttons may have been replaced at the same time. The thread attaching the tapes appears to be cotton as well and is like the other thread. Only the presence of the original buttonholes proves that the drawers had buttons when they were manufactured.

Lt. Selden’s drawers have seen significant “wear and tear” from use. The leg inseam on the right leg has either separated or been ripped apart and the stitching in the crotch has come apart around the fly. There is significant staining throughout the garment as well. While this is assumed to be from field “in service” use, it is also possible (but unlikely) that he continued to wear them after the war.

The image above shows the pattern characteristics of RD drawers seen from the front. The garment is essentially constructed of two leg panels, two waistband pieces, and two fly facings. Each leg is a single panel seamed together on the inside leg. The two leg panels are sewed together with a seam starting at the crotch in front and extending around and up the back. The waistband has one button for closure in front and is split in the back. One button closes the fly and a button at each cuff closes the leg around the wearer’s ankle.

As manufactured, the leg panels were apparently cut to only a single size pattern. The waist size was varied for waistbands of different lengths by a pleat taken in the leg panel halfway around. This adjusted the leg panels to the desired waist measurement. The location of the pleat is shown in the image. As discussed in detail later, in this way, a universal, “one size fits all” leg pattern could be utilized, saving the “cutter’s” time and effort, and implying that the cotton fabric used was plentiful enough that minimizing textile usage was not a concern.

The legs are not straight but taper from hip through the knee to the ankle.

These images compare the Hollyday and Selden drawers as viewed full length from the front. The detached right leg seam and tie tapes of the Selden example are visible.

In the image, the characteristics of RD drawers are shown as viewed from the back. The waistband, it is noted, gradually tapers front to back so that it is slightly narrower at the back than in front. The split in the back of the drawers extends down the back seam several inches below the waistband. At the end of each side’s waistband is a grommet hole for a tie string or lace to adjust the size to the individual soldier’s needs.

The leg inseams are sewn together, and the seam allowances felled down on the inside to prevent raveling, At the bottom of each leg, the seam allowance is turned under on each side and felled to provide an opening at the ankle. The end of each leg is turned up and overcast to form the cuff. Buttons and buttonholes provided allow the cuff to be closed around the ankle.

The image takes a closer look at the characteristics in the area around the waist and abdomen of the Selden drawers again looking from the front. More details of construction in this area can be seen. The waistband on each side is made of one piece of fabric folded over. The string grommets are visible as is the split extending down the back seam discussed above. There is a single button for closure on the waistband.

The fly opening on the drawers has no extension piece providing overlap as in pants. Closure is provided by a button on the right-side leg panel and buttonhole on the left. There are facings inside to reinforce this area which will be discussed in more detail later.

The images above show a comparison of this area on both the Hollyday and Selden examples. The greater length of the fly opening on the Selden drawers is noted in this comparison. The position of the button and buttonhole relative to the waistband of each appears close to the same on both. Because the stitching in the seam between the two leg panels on that pair has broken and the two sides are detached in the crotch area, the original length of the Selden example’s opening is now difficult to determine precisely, but probably the original length on that pair was roughly the same as on the Hollyday drawers.

These images show the waist and butt area in both examples viewed from the back. The split in the back seat seam is noted again, as is the tie string present on the Hollyday example. It cannot be said whether this is original to that pair as issued or was replaced. The relative fullness of the butt area in both pairs gives some indication of the fit which the pattern would have been provided for the wearer.

The above images show the waist area of both examples opened to examine the construction characteristics from the inside. As shown on each, once attacked to the leg panels, the waistband was turned inside and overcast down to secure it in place.

The fly facings and the method of their attachment is also shown in these pictures. More about the exact construction details will be discussed in the coming slides. The length of the original fly opening on the Hollyday example and the extent to which the crotch seam has been detached in the Selden drawers can be better judged from these images.

Details of the fly on the Hollyday drawers including the waistband and closure at the crotch are shown in the images above. The waistband was constructed from one piece of osnaburg, folded over. The outside attachment of the waistband on each side to the pant leg panel as shown in the closeup that also shows the waistband buttonhole. Generally, this was done by sewing the waistband to the leg at the waist, right sides together, and turning the waistband up with the seam allowances pressed up inside. More on variations associated with this construction will be discussed later.

The facings on the Hollyday drawers were attached by first sewing them to the outside leg, right sides together, then turning them over to the inside and pressing the seam flat. The fly button (Gibson “pant” size) and corresponding buttonhole for closing the fly are shown in a second closeup.

The two leg panels were connected at the crotch by a seam extending up the back. Bar tacks were added to strengthen the seam where they join. The bar tacks are shown in the last closeup.

These images show the same details on the Selden drawers and illustrate a subtle difference in how they were constructed. The inside facings on this example were attached to the leg panel along the front edge differently from the Hollyday drawers. On this pair, as seen in the closeups, the edges of the legs at the fly and the facings were simply turned under and then a line of “running’ stitches along the edge used to hold them together. This method is probably less labor intensive (quicker) than the way it was done on the Hollyday pair but also appears less “finished.” At some point, the left side seam between the waistband and fly buttonhole this stitching came undone and was repaired with “whip” stitches using different thread.

This image returns to an inside view of the Hollyday drawers to illustrate how the fly facings were positioned and attached. As is shown in the closeups, once attached along the front edge and turned inside, the rear edge of each facing was turned under. On both sides of the fly, that rearward edge was then overcast to the leg panel. The waistband was completed by turning under its inside edge and overcasting it down inside of the leg panel along the waist so that it overlapped the tops of the facings. This process was basically done the same way on both examples.

The images above provide a comparison of the facings and their construction on the two examples. The difference in how the front edge of the facings was handled is again visible but otherwise they are quite similar in pattern and execution.

As discussed earlier, one of the notable characteristics of RCB drawers, at least based upon the evidence from these two examples, was use of a “sizing” peat in each leg panel which apparently was made to a “one size fits all” pattern. The size of the finished waist on each was adjusted by using this pleat in the leg panels to match the desired waistband length. That is, the bigger the pleat the smaller the size of the drawers. This, conceptually, would allow the RCB “cutters” to work to a single leg panel pattern for all RCB drawers and for the pieceworkers to make different sized drawers by matching the leg panels to the size of a waistband supplied in the kit they received. The above image shows the large, rather “deep” pleat found in the Hollyday drawers.

Federal QM made drawers apparently were made to patterns in three discrete sizes based upon the (draft) 1865 Quartermaster’s Manual. Some contract drawers had darts for fit in the same place as this pleat in RD drawers and some (e.g., prewar Schuylkill Arsenal) were seamed along the outside of the leg instead. All also had grommets at the back for laces or string ties like the RCB examples. Other “one size fits all” patterns such as the 1864 Sanitary Commission pattern achieved fit with multiple rows of grommet holes at the back to allow adjustment by the soldier himself.

The images here show the pleats at the waistband on both examples. The Hollyday drawers have a pleat approximately 1 ½” deep while the Selden ones are about ½”, this amounts to a difference of 4 inches overall in the waist size between the two.

The back split on the Selden drawers is shown from inside in the above image and illustrates how it was constructed. The seam allowances of the back seam were turned over and felled down to a point approximately 3 inches below the waistband. At that level, the allowance is clipped on one side so that the each side edge can be turned over and felled along the split up to where the waistband joins. After the waistband was folded over it was overcast down to cover the end of this edge on each side.

The picture also shows again the placement of grommet holes for lacing the drawers in back for better fit.

A comparison of the rear split construction of both examples is shown in the above images. Interestingly, neither pair has any form of “bar tack” or reinforcement where the split begins. As this would be assumed to be a stress point, it suggests that the felled seam was considered strong enough to prevent ripping or broken stitching there in use. As indicated before, the “string tie” arrangement in the Hollyday drawers is not known to be original and could be a later replacement.

A strange “anomaly” observed in the construction of Hollyday drawers only is shown in the above images. Generally, the waistband attachment method on both examples was as described above, i.e., the outer side of the folded waistband was sewn to the leg panel (right sides together), then turned up with the seam allowance pressed inside, and finally the other side overcast down on the inside. In one area of the Hollyday drawers, however, it was sewn in place outside with visible overcast stitches. As the thread used is consistent with what was used elsewhere in construction of the drawers, it is not possible to determine now if this was a repair or represents an error rectified by the added stitching. From a functional perspective the ultimate result is the same, but the oddity of how it was done is noted.

The construction and finishing of the ankle and cuff end are the shown in the above image and associated closeups of the Hollyday drawers. The use of buttons for the closure around the ankle while seen in some civilian examples is less common that the use of tapes. All Federal examples that the author is familiar with used tie tapes as well. However, buttons may have been a favored alternative for the RCB either because of the ready availability of Gibson contract examples or the lack of ready sources for tape suitable for the purpose.

The inseams on the legs were turned and felled to prevent fraying in use. Several inches from the end of the leg the seam ended, like the back split, one side of the seam allowance was clipped and both sides felled the rest of the way down to end of the leg. Also like the back split neither a bar tack nor a reinforcement was placed at the point the opening started to prevent it coming apart or ripping there in use.

The bottom edge of the leg (cuff) was finished by turning it inside and felling it to the leg around the periphery. Finally, the buttonhole was made and stitched and the button attached to the other side completing the job.

The bottom of the legs in each of the studied examples are compared in the images shown above. The tie tapes that Selden or someone else added to replace the button closure can be seen but as discussed previously, the original buttonhole remains. While speculative, the very closed nature of this buttonhole may suggest that Selden had this work done when he first received them and not after having worn them for any length of time. Also, Selden’s pair have obviously ripped (failed) at the point where to leg seam was joined together, perhaps the result of no reinforcement at that point.

Otherwise, the similarity of the way in which that they are each constructed and finished can be seen.

This concludes the discussion and comparison of the drawers in this study. Next will be an analysis of the characteristics of the Henry Hollyday shirt which, it will be asserted, is a representative example of shirts made by the RCB and issued to soldiers in the Virginia by the C.S. AQMs.

• Men’s shirts in the mid - 19th Century

  • Pattern and style – transition era

  • Construction

• Henry Hollyday’s shirt

• A “Unicorn” – What one example can tell us

• The approach for this study

The Henry Hollyday shirt represents a unique opportunity to view another RCB product and shows yet another aspect of their operations. Certainly, as pointed out in the introduction to this part of Volume II, shirts were a significant portion of the clothing made under the RCB’s auspices and many of the seamstresses and other pieceworkers who worked for it were involved in their production. The prewar U.S. Quartermaster stated that an “industrious person by close application can make twelve [soldier’s] shirts per week.” The managers at the RCB probably expected no less and it is assumed that a significant portion of the 3000 to 4000 finished garments they received daily were shirts.

Men’s shirt styles and even the basic patterns were in transition in America at the time of the Civil War. In the 1850’s the “square cut pattern” used in both men’s work and dress shirts for at least the preceding two centuries was beginning to be supplanted by a more fitted style, originating abroad, as the shirt itself also became a more “presentable” garment for public display. Such change occurred slowly, however, especially in military clothing which largely used the traditional pattern. William L. Brown III in his excellent treatise on civilian men’s shirts (Thoughts on Men’s Shirts in America, 1750 to 1900) looks at the saga of this undergarment and its various forms as well as their role in the material culture of America in the century and a half before the beginning of the 20th Century.

Certain elements of the pattern and textiles used in making common civilian mid-19th Century men’s shirts which are discussed in Brown’s book relate directly to the Hollyday example. As a single representative of shirts made at the RCB; understanding these elements provides a context for considering its characteristics and, therefore, strongly suggests how it might relate to others made there during the war. Stated another way, even though it is a single example, to the extent it typifies or differs from the type of shirting commonly worn by men in the period, it is highly likely that all shirts made by the RCB would typify or differ from period shirts in similar ways except in minor details.

To begin this discussion, the evolution of civilian shirt designs and construction methods of the first half of the 19th Century will be reviewed and the primary characteristics of men’s “work” shirts of the pre-Civil War era highlighted. This will be followed by a detailed analysis of the Henry Hollyday shirt itself and how it was constructed. The approach to this will differ from the others in this presentation in that the analysis of pattern and construction characteristics will be focused only for this specific example and compared, somewhat abstractly, to those of typical civilian working men’s shirts in that period as well as other military shirts of the Civil War.

The above illustrations provide an abbreviated evolution story of shirt styles in the 19th Century through the Civil War era. The time intervals are very approximate. There are two reasons for this. Introduction of new styles often lagged their widespread public acceptance and older styles tended to linger in more remote or conservative populations. During most of this time, the basic pattern was the same as in the period before it started. This pattern, the “square cut pattern,” did not originate from the tailor’s “art.” It relied on simple shapes, squares, rectangles, and sometimes trapezoids with little ‘tailoring” involved. Collar size and sleeve length perhaps varied but other dimensions were determined more by the width of the fabrics and the whim of the maker. Most men’s shirts in this period were made by family members at home. While by the 1830’s and 40’s there were published “guides” for seamstresses, mostly these dealt with stylistic details not with how the basic shirt was laid out or constructed. In the figures above, the “common shirt”, “dress shirt”, and “workman’s shirt” are all “square cut pattern” shirts.

The basic elements of a “square cut pattern” shirt are illustrated in the 1840 – 1865 “workman’s shirt” illustration. The body was a large rectangle, which may or may not be seamed from two separate smaller pieces at the shoulder. The sleeves were sometimes also cut as rectangles, but later trapezoidal shapes were sometimes used. For the neck opening, a “T” shaped slit was centered in the body pointing down the front. Square pieces called “gussets” were provided for fit, larger ones where the sleeves were attached to the body and smaller ones folded to form a triangle to enlarge the opening in the neck “T”. Often reinforcing pieces were added inside the body at the arm “holes” or along the shoulder seams and also positioned at the ends of the shirt “tail” openings and sleeve openings. Cuffs of different sizes depending on style were usually made from separate rectangles folded over. At the cuffs, the sleeves were usually gathered, as they often were at the shoulder. Generally, the neck opening was gathered to fit the collar. Like cuffs, collar shapes varied with style and were either worn standing up or folded over depending on whether neckwear (stocks, cravats, or neck ties) were used.

The principal differences in period style were at the front opening. Early shirts simply reinforced or felled the edges of the slit down the center with a small pleat at the end. This was essentially for work shirts and those for dress might have ruffles added. In the first quarter of the 19th Century men’s formal outer garments encouraged exposing of more of the chest area so enhancement to that area on shirts resulted in fancy pleated inserts. Ultimately these evolved into separate “false” fronts for shirts along with detachable collars. While button closure of the front opening was done simply early in the century, by the 1840’s separate placket pieces providing buttoned closure became typical, particularly on the common everyday work shirts men wore. Dress shirts tended to be made of fine linen while “workman’s” shirts were made from coarse linen (early) and cotton. Generally, patterned fabrics were rarer than undyed or colored ones and appeared later in the period.

The evolution of men’s shirts to the modern form began in the 1850’s in Europe. More fitted styles combined with the greater social acceptance of men “displaying” their shirts in public, if not yet in “polite” social company, led to a more tailored look and changes in pattern to match. The so called “French style” shirt was the result. By the end of the war era this was seen more frequently. This change also corresponded with the widespread emergence of sources for “ready-made” clothing, supplanting “made at home” as the primary method for production.

The image shows the Henry Hollyday shirt viewed from the front; its salient features are indicated. It is a “square cut pattern” shirt made from undyed cotton osnaburg somewhat heavier than typical cotton “shirting” and much the same as was used in drawers. It has a one-piece “fall-over” collar and a separate placket sewed on the front for closing the front opening. There is one neck button and two placket buttons. The sleeve is gathered into the cuff but is not at the seam where it joins the body (more on that later). The cuff is relatively wide and, as on nearly all period shirts, the button and buttonhole are close to the wrist end of the cuff. The shirt tail raw edge is simply turned at the end and felled as it is one side of each of the tail slit. The other side edges are the fabric selvage.

All the buttons on this shirt are Gibson contract “shirt” (small) size buttons. They are attached with the original undisturbed thread like is used throughout the rest of the shirt. Like Hollyday’s drawers, their presence, taken together with the fabric from which they are made and their provenance, almost certainly identifies this shirt as from the RCB. The use of selvage edges as part of the seam allowance for the side seams on the shirt is because the chest size of the “square cut” shirt was not determined by individually fitting it but, for consistency, the width of shirt the RCB made was sized to the smallest loom from which the fabric it used was woven on. In 1864 RCB records, cotton fabrics used in production came in 27” (3/4 wide), 31½“ (7/8 wide), and 36” (4/4 wide) widths from three different mills located in Petersburg, Virginia. The Hollyday shirt probably was cut from one of the wider varieties.

Comparison with the illustrations of 19th Century men’s shirts, shows that the Hollyday example fits very nicely into what could be called a common “workman’s shirt” type from the 1840 to 1865 era. It shares significant style and construction similarities with at least three examples presented In Brown’s book that he dates to that period. Some common features are not present, however. It does not have reinforcements in the chest or back at the arm seams or at the shoulder. The tail slits are not reinforced either nor are the openings in the sleeves at the wrist. This probably speaks to the simplicity implicit in this “no frills” shirt design and that, since it was considered a replaceable issued garment, long term wearability was not as essential as in the “home-made” variety.

Perhaps because of these omissions, a post manufacture repair is also noted. At some point, possibly during the war but maybe afterward, Henry appears to have ripped it, necessitating this repair. This type of damage was probably the reason savvy mothers and wives took the time to add such reinforcements to shirts they made for their menfolk. More damage will be highlighted in the next image and details of the repair will be shown later.

Also, the “no frills” nature of the shirt is reflected by the lack of pockets. Many Civil War soldiers’ shirts do have pockets. None of the shirts illustrated in Brown’s book have them and only one shown in his period pictures does either. The reasons for this are that in the period it was expected that a vest or a coat, would be worn over the shirt and pockets would be provided those garments. Brown concludes that pockets only first appeared in “overshirts,” separate garments like smocks worn over top of regular men’s shirts for work or other reasons in lieu of a vest or coat. Some other, what are believed to be Confederate QM issue shirts (not RCB), do incorporate pockets so apparently there was no consensus among all Confederate Government manufacturers. On the other hand, the Federal domet flannel standard issue shirt had no pockets, while contract variant shirts and state issue examples often did.

This image shows the Hollyday shirt from the back. The folded “fall-over” collar can be seen in this view as well as the left sleeve shoulder seam and cuff. The lack of gathering at the shoulder end of the sleeve and gathering at the cuff on the shirt are clearly depicted. This tends to be a feature consistent with shirts made later in the pre-Civil War era. It also was a pattern feature that reduced work for the assembler. Federal standard issue domet shirts were made this way and most even had cuffs without gathering further simplifying the construction. In those cases, the separate cuff was sometimes eliminated altogether, replaced with an inside facing piece attached to the end of the sleeve. Also visible in this view is the back side of the tail.

Further evidence of damage to the shirt is clearly seen along with the repairs it necessitated. This area will be more thoroughly inspected later in the presentation, but it is unclear whether these happened in a single incident or at two separate times. None of the repairs are believed to be professional conservation but look to be amateur work. No professional conservation has been done either by the current owner or recorded in its previous history.

These two images provided by researcher Ross Kimmel, give an excellent idea of the how the arm gussets are placed in the construction of this shirt. The picture on the right also shows the degree of gathering of the sleeve done into the cuff. This was obviously a rectangular sleeve pattern. The sleeve opening slit and the cuff buttonhole are noted as well.

The neck gussets are also visible. In some period civilian examples, there is a reinforcement added to this area running from the neck to the shoulder seam which covers the gusset and, if the body is seamed together, that seam as well. This added feature was not included in the “no frills” construction of the Hollyday shirt.

The images above show details of the “fall-over” style collar on the Hollyday shirt. The pattern’s design naturally allows the collar to lay over smoothly when folded down with the base acting like a separate neck band. If worn with a cravat it could also remain standing up to accommodate such neck wear. This design is quite common in pre-Civil War era civilian men’s shirts. Interestingly, U.S. Army issue shirts dating back at least to the War of 1812 eschew this design and utilize rectangular collars which work well when a neck stock or wide cravat are used, but do not fall over smoothly when folded down. The standard issue Federal domet shirt of the Civil War retains that characteristic.

The gathering at the neck to fit the collar the neck opening with the gussets in place is also seen clearly in these images. Again, such gathering is usual in pre-war civilian shirts although pleats, particularly at the back, are also seen. Late in the period, oval cutouts sized to the desired neck opening were also used. Most home-made work shirts, however, used the traditional “T” neck with inset gussets that required adjustments (gathering or pleats) to fit the collar. The standard issue Federal domet shirt also was constructed this way as US Army shirts had been for over a half century before the war.

The neck button on the collar is a Gibson “shirt” size button. It is attached with undisturbed original thread but, from the condition of the button’s surface, the shirt was soaked with water, possibly laundered, at least once but more probably multiple times in its history. The striations visible in the closeup picture of the neck button are due to this. The other wooden buttons seen on the shirt also show damage due to the same type of exposure. In his book, Brown points out that early period civilian shirts (circa 1760 to 1800) used “dorset” buttons at the collar which were constructed entirely from thread and removable “sleeve links” at the cuffs so that the shirts could be washed easily. Later in the period, civilian shirts used bone, milk grass or ‘mother of pearl’ buttons not effected by water for the same reason. Further evidence that the RCB issue shirts and drawers were probably considered ‘expendable’ in some sense was the use of these wooden buttons which would be so affected by laundering them. Federal undergarments used metal buttons which also were less tolerant to multiple wash cycles leaving rust stains.

Details of the placket construction on the Hollyday shirt are displayed in these images. The placket was constructed from a separate piece of shirt fabric sewn to the front of the shirt at the split. It has been folded over along both edges and secured with back stitching. Close inspection shows the meticulous care with which the maker executed this stitching, such that from the front it almost appears to have been done on a sewing machine. This stitching continues around the bottom end which was turned up and across the placket slightly higher up to reinforce it at the opening. A pleat in the shirt below the placket allows for overlap to accommodate attachment of the placket and the slit edge on its right side, overlapped by the placket, which was turned under and stitched down.

Two buttonholes have been neatly worked into the placket, evenly spaced with the neck button to provide closure. Gibson buttons are attached to the right side where it was turned under. These buttons exhibit the same surface roughness described above suspected to be through exposure to water possibly from laundering.

The construction of the placket on this shirt is directly analogous to the method used on the three 1840 to 1865 era “workmen’s shirts” discussed in Bill Brown’s book. It is the standard way in which plackets were attached throughout this period. As early as the first decades of the 19th Century, the simple neck slits on some common men’s shirts were closed with buttons. In that early era the edges of the slit were sometimes simply turned under to provide reinforcement for the buttonholes and attachment of buttons. In other examples, such reinforcement was provided by a second piece of fabric added inside the shirt along the slit. As in this shirt, gathering or pleats in the shirt body were used to compensate for overlap when the shirt was buttoned. As has been already noted, the standard issue Federal domet shirt did not have a placket or button closure. Some Union contract variant and state issue shirts were made with placket construction like the Hollyday example, adhering to the standard style of the period.

Characteristic details of sleeve construction on the Hollyday shirt are illustrated in the images above. The shirt, however, is odd in that there is a significant difference between the overall construction between the right and left sleeves. This will be discussed next but first, what is common, will be presented.

As shown, there was no general gathering of the sleeve at the shoulder end. The sleeve is sewn to the body and the seam allowances turned under and felled on the inside. At the cuff end the sleeve is gathered to match the cuff piece, which was made from a single piece of fabric, folder over. The gathering is tight, uniform, and well executed indicating the hand of an experienced “seamstress.” The wrist slit was turned under and felled on each side before being joined at the lower sleeve seam which, like the one at the shoulder, had the seam allowance inside turned under and felled along the rest of the sleeve. No reinforcement or “bar tack” is provided at the end of this slit.

A very neatly done buttonhole and a Gibson button were added at the wrist end of the cuff width, placement consistent with the style of the period.

The images above show very odd variations in sleeve construction for a single example which are hard to rationalize. At a primary construction level, the two sleeves on this shirt were each made differently. One (the left) is made of two pieces of shirt fabric that have been seamed together while the other (the right) is made from a single piece of shirt fabric which seems to have been slightly too large, at least, at the shoulder end. For the left sleeve, the two pieces were sewn together along what is the top of the sleeve to make it full width and the seam allowances turned under and felled as elsewhere in the shirt’s construction. The right sleeve is one piece, not seamed together, but while not gathered at the shoulder, it has a single pleat taken right in the center that is somewhat crudely basted down. It is possible that this pleat was done to make up for it being wider than the left sleeve so that once attached both sides would appear the same. Gathering at the cuff would have resolved any small difference between the two there. The amount taken up in the right sleeve pleat was approximately the same as seam allowances on the left side. This suggests that the two pieces used to make the left arm might originally have been cut one half the full width of the right, but this was not verified.

The resulting questions are, how did this happen and what was “normal” in examples made by the RCB? Obviously, the true answer to the first is unknowable but several options are possible. It is conceivable, but unlikely, that both sleeves were one piece in the kit and for some reason the assembler cut one of them in half before realizing the mistake. The cut sleeve then had to be sewed back together to make the shirt and a pleat was taken in the single sleeve piece so its width would match the joined sleeve. Alternately, it could also be that the sleeves were supposed to come in two pieces and the “cutter” at the RCB neglected to completely cut the one sleeve to separate it properly resulting in the pleat. Also, unusual, but perhaps somewhat more probable that the first. Finally, it could be that during “cutting” at the RCB, to better utilize scrap material, one sleeve was intentionally cut in two separate pieces leaving the “pieceworker” with the job of making everything work out. We will never know for sure but however it occurred clearly one side was made to one-piece sleeve pattern and the other to a two-piece pattern. So, which was the actual RCB pattern?

In terms of how it was supposed to be made, the sleeves in the “square cut pattern” shirts for over two centuries prior to this shirt were essentially made the same way. They were single piece, rectangles sewn together on one side to where they joined the arm gusset. This is not to say that there never was a shirt made using this pattern where the sleeves were made up from two pieces joined together, but that would have been very unusual. Furthermore, it is doubtful the RCB would have patterned its shirts in this way. Workwoman’s Guides for shirts from the mid-19th century provided layout arrangements to maximize the utilization of fabric of various widths. As discussed in Part 2 relating to the pant examples with two-piece pocket bags, cotton goods, especially the heavy osnaburg used in so many RCB garments, does not seem to have been in limited supply except perhaps very early in the conflict and, maybe, for short intervals at other points during the war. The RCB’s supply of such textiles came from three mills in Petersburg so resupply should normally have been possible very quickly. Husbanding the use of this fabric would hardly have seemed necessary. Using a two-piece sleeve pattern on shirts to better utilize “cutting table” scrap, while possible, does not sound consistent given the extra labor (and hence, lowered unit productivity) which would result. Again, with only a single example to judge from, absolute certainty on this point is impossible, but it is strongly believed the shirts normally were made with one-piece sleeves.

The arm gusset arrangement and construction can be seen in the above images. Normally the gusset is attached to the sleeve before the sleeve is sewn to the shirt body. As with most other seams throughout the shirt, the seam allowances are turned under and felled, that is, overcast to prevent fraying through use or in laundering. The main seams appear to have all been done with running stitches. The form of the overcast stitching is also called a “hemming” stitch. It is tightly controlled with respect to the edge which is being felled so that on the opposite side it looks more like a “running” stitch that typical overcast (“whip”) stitches. Throughout this shirt the overcast work is done in such a competent hand that it is extremely hard to tell the difference looking from the outside of the shirt. This stitching is visible in the enlargement of the gusset.

The small gussets in the neck of a “square cut pattern” shirt are significant features often not visible in pictures. Their construction is usually obscured because of reinforcements sewn onto the shirt body which cover them. Reinforcement is needed in this area because the form of the neck “slit” combined with the shape of the gusset itself concentrates the stresses from the wearer’s movements at that point in the shirt body leading the ripping. The “no frills” approach to this shirt’s basic implementation and resulting damages to this shirt in that area make this point! It is noted that on the Federal standard issue domet shirt, narrow rectangular reinforcements were provided to reduce the stress at the neck even though in some ways its pattern implementation is even a more basic, even “retro,” version of the same “square cut” design.

The above pictures show the neck gusset on the Hollyday shirt both from the outside of the shirt and as it appears from the inside. The gusset originally appears to have been a square piece of shirt fabric which was doubled over to form a 45-degree right triangle. The edges were turned under to eliminate raw edges and pressed. The gusset was sewed into the neck opening at the shoulder seam to enlarge it. As can be seen in the images, the gusset and neck opening in the shirt’s body were gathered to match the length of the collar before it was attached.

The damage to the shirt and the repairs has been referenced several times in the above discussion. It is not known when the damage occurred nor if it all occurred at the same time. Given these unknowns, the question of whether these are field repairs made during the war, home repairs afterward, or later amateur “conservation” cannot be answered with certainty.

The above images show damage and repairs in the center back of the shirt. There is a large “hole” underneath the collar near the middle of the back and a rip radiating from it. It appears a piece of fabric has been removed (torn out) from the back. This original repaired area is seen in the closeup. The remnants of thread around the “hole” may have been an attempt to repair the area at some point. Stitches are also visible, made with white thread, used to sew together the sides of the rip emanating from the “hole.” All the stitching is fairly crude perhaps suggesting a “quick” repair made under adverse (field?) conditions but just as likely they could just have been done by someone who was not skilled in their needlework.

Details of the other repaired area to the Hollyday shirt are seen in these images. The rips in both the front and back of the shirt emanate from the high stress point at the apex of the neck gusset where it meets the top shoulder seam. This certainly suggests that these both may have occurred at the same time. The felling of the shoulder seam would have provided some reinforcement to prevent the seam’s stitching from breaking but the unreinforced osnaburg fabric in that area was ripped by whatever the force was exerted on that area. Both sides were repaired in same way as the rip in the center back using similar white thread.

Did all this damage occur at the same time? It simply cannot be said for sure, but the causes seem to have been different and under close inspection, the stitching seen above on this damage seems to be executed in a slightly more competent hand than that in the back. This would possibly imply the damage did not happen at the same time but is certainly not conclusive.

It is also difficult to say whether the damage/repairs happened during Hollyday’s wartime service. None of the repairs are the work of professional conservators but without doing thread analysis, it cannot be known if these repairs are contemporaneous with the war. Pictures taken in the 1980’s, before the shirt was acquired by its present owner, do show the repairs pretty much as they appear today so the work is known to be at least 30 years old.

The construction details associated with the shirt tail on the shirt are shown above. The side stitching of the shirt’s body runs from the bottom of the arm gusset on each side to the beginning of the tail slit. One edge of the body pieces on each side is selvage and the other raw indicating that the pieces were made from osnaburg fabric wider that the shirt width. As was indicated earlier, cotton textiles procured for garments by the RCB were in 3/4, 7/8, and 4/4 widths suggesting that this shirt pattern was probably sized to be constructed from 3/4 wide fabric. In this case, one of the two wider varieties was likely used and the body piece needed to be trimmed in width. The body panels were positioned such that one selvage edge is on each side of the shirt. Where a selvage edge was present felling was not required so the raw edge at edge side seam was simply turned under and overcast. The bottom edges were handled in the same manner.

In period civilian shirts there was usually a reinforcement of some kind inserted at the tail split to prevent the seam from separating or ripping at that point. These are not present on the Hollyday shirt, Even the precautionary step of making “bar tacks” there was not done in the “no frills” construction. Federal standard issue domet shirts typically did reinforce the shirt in this area.

The stitching on the shirt tail is seen from the inside in the above images. The way in which the felling of the raw edges is done, and the precise execution of the overcast stitching are visible in the closeups. The seamstresses who made these shirts clearly show amazing “handwork” even when working quickly. While in several places of this discussion, the “no frills” nature of the pattern and construction of the Henry Hollyday shirt has been emphasized, the competent way in which it was executed was neither haphazard nor sloppy.

This concludes the description and analysis of the Henry Hollyday shirt, its characteristics, and anomalies.

• What can be said based upon a single example?

  • Consistent with civilian examples circa 1845 – 1865

  • Fabric, notions consistent with that used in drawers and uniform linings

  • Simply but competently constructed – “no frills”

• Implications on understanding other “RCB made” shirts during the war

  • Typical or not?

  • The RCB production model?

• Other types of shirts (e.g. flannel) and early war production

  • Other C. S. Quartermaster Issued Shirts

In the previous section, characteristics of Hollyday shirt were presented, but the question of just how representative this single example is of all RCB made shirts over the course of the war was not directly addressed. Even the proposition that it was produced by the RCB is only based upon strong, but indirect facts like the presence of Gibson buttons, similarity of the shirt fabric to that used in other garments attributed to the RCB, and a history that links it to the principal theater of the war which that clothing bureau supported. What can be said definitively is that the Hollyday shirt is a simply made, “no frills” shirt made from commonly used fabric. The case that it represents all RCB shirts is supported by its close similarity with what was the “typical” civilian “workman’s shirt” of the period leading to the war. As such, it is archetypical of what might be expected to have been made for enlisted men at that time. While this is a useful way of framing the question, it still leaves a lot of unknowns which are unanswerable without other examples for comparison.

Some attempt at other relevant comparisons, for example to Federal issue shirts, were made as well. It is perhaps significant that this Confederate example adheres to the same “no frills” philosophy as the Federal standard issue domet shirt and that Union variant issue shirts were, for the most part, even closer in pattern and construction to the Hollyday example. However, this is still peripheral to the question of how representative this shirt is of other RCB shirts.

If the production model for the Federal Schuylkill Arsenal production facility and what we know about the RCB clothing manufacturing operation which followed the same model are considered, some further insight is derived. Documentation of Schuylkill Arsenal’s operations going back at least until the 1820’s indicates that soldier’s shirts there were cut out to a specific pattern and made up by local pieceworkers in roughly the same process used in the 1860’s. Furthermore, based upon research published in 1993 by Steven E. Osman, the pattern of shirts at both ends of that decades long time window were essentially the same varying only in the fabric used and some minor construction details. At the RCB while details of RD jackets did change, the basic pattern remained invariant during the entire production period. From the study of RD pants in Part 2, these were also made to a consistent pattern.

The point is that both QM Departments were inclined to a substantial degree of inertia when it came to changing patterns unless compelled to do so for other reasons. When combined with the Hollyday shirt’s consistency with comparable civilian shirts and Federal issue examples, this all strongly suggests that the shirt should be representative of other RCB cotton shirts made during the war. In final analysis, it is asserted that the Henry Hollyday shirt probably exemplifies the cotton shirts produced at the RCB during the war, at least, until different, identifiable examples surface. It is hoped research can uncover such additional examples that can clarify this assertion.

The QM Department in Richmond also issued other shirts besides RCB made cotton examples. Documentation of field issues records numbers of either all wool or (more likely) wool – on – cotton flannel shirts were also distributed. In late 1863 to 1865 the Shipping Book lists large amounts of English flannel received at the RCB. Some of this was obviously made into shirts by the RCB but whether the pattern and construction of those is like the Hollyday example is unknown. Numbers of English made shirts (both flannel and cotton) are also documented to have been received by the Richmond Depot through the Blockade for issue to the army.

Given the last discussion, one may ask was there a “standard issue” shirt for soldiers across the Confederacy? While no catalog of possible C.S. Quartermaster issue shirts will be attempted in this presentation, the above images show several which illustrate the diversity provided to soldiers. Like most shirts with provenance suggesting QM production, the assertion that they represent “standard issue” examples is more by inference than based upon solid evidence. The author has not had the opportunity to personally inspect any of these shirts, so the observations made a solely based on pictorial evidence.

The Port Hudson example is said to have been taken by a Union soldier from Confederate stores following its fall in mid-1863 after the surrender of Vicksburg. The Andrew Thomas Beam shirt in the collection of the American Civil War Museum in Richmond belonged to a South Carolina soldier killed at the Battle of the Crater in Petersburg. The John Jay shirt belonged to a member of the 1st Louisiana Heavy Artillery who was part of the garrison at Vicksburg until is death of sickness December 1862. The location of the shirt and a pair of pants he owned is currently unknown (Reference Fred Adolphus, "Confederate Uniforms of the Lower South, Part II: Tennessee, East Louisiana, Mississippi and Alabama”, 11 August 2019, adolphusconfederateuniforms.com). The Lt. John Selden shirt is included not as a C.S. QM made shirt at all, but an example of British shirts imported through the blockade in large numbers throughout the war for issue to troops in the Confederate service. It is marked with the British government’s “broad arrow” and dated 1859. Selden could either have received the shirt from the field AQM serving his unit or purchased it from Ordinance Department stores as described in Part 2.

Based on the limited information available, Beam, who was a private in the 23rd South Carolina Infantry, probably received that shirt before his unit was transferred to Virginia from South Carolina in the May 1864 timeframe several weeks preceding his death in July 1864. Before their posting in South Carolina earlier in the year, the unit had served in the Department of East Louisiana, Mississippi, and Alabama, so he may also have received it there. Its similarity to Louisiana Private John Jay’s shirt suggests that the two may, in fact, be related. The major difference noted from the images is the presence of the exterior breast pocket on the Beam example. The Port Hudson shirt was sold through a private auction house a number of years ago and now (I assume) is in a private collection. It also has a pocket but a different collar pattern. Unfortunately, from this picture it isn't clear if the shirt’s pattern also includes a yoke reinforcement like the Beam and Jay examples or whether what is visible is simply a shoulder reinforcement strip. Since the specific source of manufacture in all three cases is speculative, these variations could indicate they were made by three different Depot/Bureaus but, of course, there is no way to know. While conceptually similar to the Hollyday shirt these three exhibit differences in pattern suggesting, like Confederate “Depot” jackets, the shirt designs varied from region to region and were not standardized.

As a British product Selden’s shirt was made to a traditional “square cut pattern” but its implementation is fundamentally different from either the Hollyday example or the Federal standard issue domet shirt in that the body fabric is folded along one side and seamed on the other with holes left for the sleeves and tail slits. This apparently was also an early variation on the “square cut” shirts as one documented in Bill Brown’s book dating in the second half of the 18th Century is made the same way. Such British army shirts have long referred to as “Ammunition shirts.”

This concludes Volume II and my study of Richmond Depot Clothing. Before summarizing my thoughts, I would like to go back and reflect on what is one of my most vivid impressions after nearly eight years of investigation and study of these artifacts. From the very first jacket I photographed through the items presented in this volume, the “Art” of the seamstresses and pieceworkers who assembled them has been notable. Generally, this “handwork” of common women done 160 years ago has been underrepresented in other research studies and needs to be highlighted in its own right as a significant “characteristic” of the garments made by the Richmond Clothing Bureau and its employees.

After that discussion, there will be a summary of my thoughts from a material culture perspective concerning what I have seen and learned in this “journey” into the clothing produced by the RCB.

This montage of images presents details of the needlework exhibited in some of the RCB garments studied in this volume. Each illustrates the character of the sewing and the range of stitches employed in their construction. It is important to recognize that professionally made clothing of the first half of the 19th Century usually exhibited incredibly fine, well executed handwork. Some elements like buttonholes were even the province of specialists who did nothing else. The precise backstitching in the unusual cuff construction on the Tinges jacket is illustrated in comparison to the well-executed professional topstitching and buttonhole on the pocket of the Selden pants to illustrate not only the precision of the professional’s work but the competence of some RCB pieceworkers in doing theirs.

The meticulously made buttonholes and construction stitches (backstitching) on placket of the Hollyday shirt is shown from both sides. Buttonholes in the fly of the Hollyday pants are a counterpoint to those of the shirt but nevertheless were executed in a neat and competent manner. Likewise, the exterior view of one of the felled seams in the shirt shows the precise, competent hand of an experienced seamstress. This can also be observed in the images of interior details showing how the seams in the Selden and Hollyday drawers were felled with overcast (or hem) stitching.

Still another case of efficient and competent work is displayed at the hem of the Tinges jacket where the lining is secured to the jacket body. Very neat overcast stitching sets the lining to the body and topstitching (seen from the back) applied to strengthen the connection.

Taken together, these images are but a few examples that display an underappreciated aspect of the clothing of the RCB, the contribution in its construction of thousands of individual workers who labored to produce it. Anyone who has ever tried to simulate their work can appreciate the skill which is evident in its execution. Regardless of the subtilties of the pattern followed or textiles used in the soldier’s clothing, this “handwork” is a testament to the (mostly) women who made it.

The first jacket I ever studied was the “Rebel jacket” that Captain Fred Sanborn of the 5th Maine Volunteers obtained from a Confederate prisoner at the Battle of Spotsylvania in May 1864. This remarkable specimen currently resides in the Fifth Maine Regiment Museum on Peake’s Island Maine. While the condition of this example is incredible and it is an archetypal example of a RD type II jacket of the first half of 1864, it is also a “tour de force” in the “Art” of the seamstresses who assembled the clothing. The stitching technique is everything described above relative to the qualities of handwork executed by many RCB pieceworkers, quickly done, but competently executed, and precise.

The image above is of the left inside of the jacket’s lining and the finish work done there. The associated closeup images provide views of the character and quality of this work. The buttonholes are executed masterfully approaching almost the levels of professional “buttonholers.” These are shown both from the front (good) side and from the back (normally unseen) side. Also visible in these closeups are the neat, precise “running” stitches used in topstitching around the jacket front and hem as seen from the back (or “bad”) side.

An interesting detail found on most RD jackets is the use of ‘buttonhole” stitching to bind the opening where the interior pocket is inserted in the left breast lining. As shown, this is executed competently but as a construction element was not worked with the same precision as the front buttonholes which are visible. It was unnecessary here because this work served a different purpose.

The stitching securing the beltloops to the jacket’s back is highlighted in another closeup. This remarkable work is unique among the RD jackets I have studied, for the needlework “whimsy” displayed by this particular worker in doing the task. Normally, crude stitches are used with the only purpose being to firmly attach the loops and finish the job as quickly as possible. This shows perhaps pride in the result or maybe even an attempt to have a little fun while doing it. Quite remarkable!

What are “material culture studies?” A dictionary definition of material culture and its study is (from Wikipedia):

“The scholarly analysis of material culture, which can include both human made and natural or altered objects, is called material culture studies. It is an interdisciplinary field and methodology that tells of the relationships between people and their things: the making, history, preservation, and interpretation of objects. It draws on both theory and practice from the social sciences and humanities such as art history, archaeology, anthropology, history, historic preservation, folklore, archival science, literary criticism, and museum studies, among others.”

A simpler, less academic way of saying this is studying the context of a man-made object. That is, when studying an artifact, the first question one should ask is: “what is its context?” Who owned and used it and why? Who made it and how? What was it made from and when? What distinguishes it from other similar artifacts? What effect did the society (culture) in which it was made have on its form or how it was used? What effect did it have on the society (culture) that made it? What was its value (intrinsically and monetarily) then and what is it now?

Since publication of Volume I (2018), my personal study of garments produced by the Richmond Clothing Bureau has expanded to include seventeen different uniform jackets, six pairs of pants, two pairs of drawers and a shirt attributable to that operation, which have each been photographed, analyzed, and, in most cases, compared. It is estimated these items were made as early as the third quarter of 1862 through as late as the first quarter of 1865. Some came with provenance that associated them with specific soldiers and others with battles or events during the conflict. As a single physical artifact, each tells a story associated with its own history and the context in which it was used. However, from this group of artifacts, it can be said that, collectively, they are representatives of the material culture of both the place and era from which they came. In reflecting on the journey that I have taken seeking out, studying, and researching these historic artifacts, I believe that what I have learned is best expressed through a material culture paradigm.

How these pieces were produced is significant, as is the context in which that production occurred. The RCB started manufacturing uniforms and other clothing for the Confederate Army in the city of Richmond in the late second or early third quarter of 1861. Its production model was based upon a pre-war U.S. Army model used by the Schuylkill Arsenal in Philadelphia for over fifty years prior to the start of the Civil War. The organizers and future managers of the RCB came from related civilian enterprises (tailors, dry goods merchants, etc.) not from a military tradition. The supply chain needed for it to operate, basically was derived from peacetime commercial enterprises. Most important, the largest part of its workforce were residents (largely women) of the Richmond community. At the start of the war this workforce, unlike that at the Schuylkill Arsenal, came to the job without experience in making military clothing. Many were work-at-home seamstresses who had made garments for their family and relatives. They were motivated, at least in part, by patriotism. By the end of the war, most of these workers were female relatives of soldiers, often in active service, disabled, or deceased, who were relying on their handwork to make a subsistence wage, very much like their “sisters” employed at the same time by Union Quartermasters in Philadelphia.

Throughout the years of the conflict, the RCB grew and matured into an operation providing serviceable clothing for the Army. The design of that clothing expressed elements of period military and civilian style. It both reflected and influenced the types of textiles and other supplies needed to make the garments. The clothing production impacted the Richmond community by employing thousands of individuals involved in making needed supplies, product assembly, and distribution of the RCB’s output to the army. Ultimately, the uniforms it produced were in a sense emblematic of Lee’s Army of Northern Virginia being widely issued throughout the war to most of its enlisted men.

It was through systematic study of this group of examples, but mainly from the uniform jackets because of the relatively large number of examples analyzed, that the most insight was gained. As a researcher, I was consistently struck by how similar the jackets were but, at the same time, how each was individually different, very much mirroring both the people who made them and the soldiers who wore them. Uniform production was anything but “mass production” in the 20th/21st Century sense but, at the same time, hundreds of thousands of jackets, pants, drawers, and shirts were produced for issue by Assistant Quartermasters in the field. The jackets and pants were made of woolen fabrics woven from local sheep’s wool in Virginia mills as well as from textiles produced in England that were “run” through the blockade. The shirts, drawers, and linings were made of southern cotton woven mainly in Petersburg, on the Appomattox River 21 miles south of Richmond and the RCB used millions of wooden buttons made in the carpentry shop of two Scottish immigrants just a few blocks up Cary street from the Clothing Bureau’s location in the heart of the city.

The differences in the jackets and pants, “anomalies” as I have termed them, highlighted in this analysis, are, I believe, largely representative of the individual background and capabilities of the pieceworkers who assembled them, so called “seamstress choice” variations. Sometimes these variations were partially in response to the character of the textiles they were working with, were enabled by the pattern pieces supplied for them to use or were in response to abnormalities in the pre–packaged “kits” they received for assembling the garments. Mainly, however, they resulted from the individuality of the pieceworkers themselves. In some cases, post issue modifications were also encountered where the individuality of the owner was displayed with the addition of a second pocket, tailoring for better fit, or even substitution of a warmer lining.

Considering the jackets, my objective was originally to answer the question of what details made a “Richmond jacket” distinguishable as such. All the examples studied answered that question to some degree, but the group fell short of providing a definitive new set of absolute criteria beyond the original ones first described by Jensen almost 35 years ago. All RD jackets used the same six panel “wrapper” pattern for the body pieces derived from mid-19th century tailored civilian dress and frock coats. They used a two-piece sleeve pattern derived from the same sources, styled with a moderately wide sleeve at the elbow and a narrow, non-opening cuff. Of all Confederate “shell” jackets, they are among those that exhibit a reasonably stylish cut from a period perspective. A similar story appears to also be true with the more limited set of pants studied in this volume.

Unfortunately, the pattern itself is not the whole story. The pattern characteristics of the jackets, pants, drawers, and shirts made by the Richmond Clothing Bureau were developed by the civilian tailors and dry goods merchants who first managed it. Pattern creation in the middle of the 19th Century for civilian coats and trousers was done in an almost mechanical way, working from standard torso measurements, and using very nearly mathematical layout methods. Final fit and stylish appearance were in the realm of the tailor’s “art,” so patterns were usually created for each customer individually. The same basic jacket and pants patterns were almost universally used in the preceding decades. As a result, many other Civil War military jackets, both Confederate and Union, were based upon the same conceptual pattern that the RCB used. The patterns for shirts and drawers as undergarments enjoyed far less stylistic scrutiny. In fact, the same basic shirt pattern had probably been used for several centuries before the Civil War began and was only beginning to change significantly in the civilian world as it ended.

Implementation of the basic jacket pattern from an assembly perspective is evidenced through steps like setting the lining or sleeves into the body or how details like collars, epaulets, facings, or pockets were constructed. Consistency in these came closer to distinguishing RCB made jackets, in that, nearly all jackets and pants studied were constructed in much the same way. For reasons that are not completely understood, however, many exhibited variations (“anomalies”) in the way some details were executed. Attempts were made in the descriptions to explain or rationalize these but remain speculation. In fact, some details like the number of buttons on the front varied significantly enough from the standard that they become almost inconsequential in attribution of a jacket to the RCB. There is no explanation for this either.

The construction techniques and handwork execution in almost all the clothing studied reflects what can be termed a “quick but efficient and competent” philosophy. This is very noticeable when compared to professionally tailored garments of the period. In some cases, care was taken to provide extra stitching or reinforcements such the inclusion of “bar tacks” at stress points. In other cases, construction details that were extraneous to what was needed for basic assembly were eschewed. Depending on the relative competence of the pieceworker in specific procedures, like finishing buttonholes, the quality of the workmanship varied significantly as well. Interestingly, some of the more meticulous “handwork” observed appears on the two pairs of drawers and the shirt studied despite their being the simplest in pattern and execution.

A surprising number of the jacket examples exhibited some amount of machine stitching in their construction. The reason this was surprising is that, heretofore, the general assumption made by researchers was that clothing produced by the RCB was entirely hand sewn. Part of this may come from the fact that Schuylkill Arsenal did not allow its pieceworkers to use sewing machines in construction of its garments. Also, the assumed nature of the workforce employed by the RCB suggested that the cost of such technology would be prohibitive for them either to obtain or maintain. While it was speculated that machine sewing might indicate small tailoring shops or other commercial enterprises participated in uniform assembly, the presence of machine sewing in statistically significant numbers (three) in the jacket sample studied (seventeen) would tend to mitigate this as the only explanation.

The textiles used in RCB garments have been another focus throughout this study. Understanding these as well as their sources provides insight into the overall supply chain that supported the clothing manufacturing operation. The fabrics used also are integral to dating the artifacts themselves. Early in the war, when the RCB operations were formulative, a great deal of diversity occurred in the fabrics procured for its production. By the middle of 1862, the QM operation was standardizing its domestic textile sources reducing this. As the war progressed, external conditions affecting raw materials availability and the incursion of Federal forces in Virginia which closed or destroyed regional mills impacted the supply of domestic uniform fabric. This resulted in increased reliance upon English woolen goods and more Quartermaster control of cargoes aboard blockade runners the last eighteen months of the war. However, the presence of “English goods” in a number of mid-war examples shows that they were also part of the mix used throughout the conflict.

A lot of emphasis has also been given to the relevance of the wooden buttons John and George Gibson produced for the RCB between November 1862 and January 1865. Certainly, their presence on the Taylor pants and Henry Hollyday’s drawers and shirt is seminal in their attribution to the RCB. Throughout the war an important part of the supply chain involved procurement of these simple components required for garment production. Large numbers of bone buttons were made for them by the Union Manufacturing Company of Richmond and shipments of English metal buttons (military and common shirt and pant sizes) received through the blockade also were significant in the production of garments. Sewing thread was purchased locally from Richmond dry goods concerns and importers but also came in from England.

Based upon the many pictures taken of “proud privates” wearing their new RCB made jackets, the soldiers took pride in their new government “duds” as symbols of their service and patriotism. Private Charles Tinges paid $14 for jackets in 1864, $12 for a pair of pants and $ 3 for some drawers, taking another jacket home at the end of the war as a memento. Had he acquired one in 1864, a cotton shirt would have cost him $ 3 as well. Each enlisted man was given an $85 allowance for the clothing that he was supposed to receive each year. That same private received $ 11 per month in wages at the beginning of the war, but after June 1864 that was raised to $18 so Tinges’ jackets cost almost a month’s salary.

In service, clothing was an expendable. Replenishment of these items was obviously necessary to success of Lee’s field army. In many communications during the conflict, he lamented the need for increased supplies to clothe his “needy” troops. However, some enlisted men avoided replacing worn items so they could, instead, receive the equivalent cash in their pay. Some even sold the clothing they received. J. B. Jones (“A Rebel War Clerk’s Diary”) records that in November 1864 he bought two shirts “of a woman in Richmond who gets her supplies from passing soldiers…I paid $15 each; the price for new ones, of inferior quality, is $50 a peace.” In a February 1865 Confederate Congressional report, Quartermaster General Lawton comments on “the soldiers’ improvidence in taking care of their clothing and bartering it off to civilians, traders, and others.” The report concludes that because of such behavior “large quantities of government clothing [is] possessed by persons in civil life and dealers in such articles.” Today, these items are worth many thousands of times what Tinges and others were charged for them then. More to the point, today they also provide an irreplaceable “window” into the life and times of the people who used and produced them.

The impact upon the population and commerce of Richmond from production of uniforms at the RCB for the Confederate Quartermaster Department was clearly significant. The importance of their products to sustaining Confederate forces in the theater of war in Virginia was equally so. After October 1862, when the act establishing the commutation system for soldiers’ clothing expenses was repealed, the Quartermaster Department assumed responsibility for provisioning the Confederate Provisional Army and RCB manufactured clothing became the standard issue provided to troops operating in Virginia. Rough estimates of between three hundred and four hundred thousand uniform jackets and as many as five hundred thousand pants, shirts, and drawers were produced for the army during the war by this operation on Cary Street in downtown Richmond and the upwards of 4000 “pieceworkers” they employed.

Dick Milstead
2021 Annapolis, MD

Dick grew up in Washington D.C. and Silver Spring, Maryland. After graduating from the University of Maryland in Mechanical Engineering in 1968, he received a SM from MIT in 1969, and a PhD. from New York University in 1975. Over his 38-year career, he worked in mechanical and electronic design, systems engineering, supply chain analysis, and capital planning/management for Bell Laboratories, Lucent Technologies, and AT&T before retiring in 2006.

Dick’s interest in American military material culture dates from the early 60’s. He was active in Living History groups during the Civil War centennial and the bicentennial of the American Revolution. In 1965, he was one of the founding members of the 1st Maryland Regiment (Revolutionary War). As part of his involvement in the hobby, he has been reproducing historical military uniforms and accouterments for nearly six decades.

A member of the Company of Military Historians for over 55 years, he has published five articles in the Company’s Journal, Military Collector & Historian, on Civil War uniforms and has been an active researcher supporting Liberty Rifles since 2017. Currently he is involved in research into Civil War enlisted uniform production in both Richmond (CS) and Philadelphia (US).

He and Sandy, his partner for the last 49 years, currently split their year residing in Rockport, Maine (summer and fall) and Annapolis, Maryland (winter and spring). They have one daughter, Kate, who lives in Boone, North Carolina.