Article

Shirts of Runaways Advertised in Rhode Island by Paul Dickfoss

“white Flannel Shirt,” “striped Linen Shirt,” “tow cloth Shirt” :Shirts of Runaways Advertised in Rhode Island

By Paul Dickfoss
Dubois’ coy, 3rd New York Regiment

Introduction

Shirts are worn by all males yet few authors have attempted to determine what fabric and colors are common, rare or unusual for these garments. Excellent period shirt patterns are available, but all recommend linen, cotton or linen/cotton blends for fabric.1 Were other fibers commonly used to make shirts? Most commercial shirt patterns recommend white, off white or check linens. One pattern is very specific in recommending unbleached linens and stripes but not checks for French Canadians, while other patterns make no mention of stripes at all. Were stripes peculiar to the shirts of French Canadi­ans? There is insufficient information to draw such conclusions.

This detailed study looks at all shirts, totaling 250, described in advertisements for civilian run­aways in Rhode Island newspapers from 1760 to 1783.2 It is intended as a guide for reenactors and interpreters when selecting fabric, color, construc­tion, buttons, and repairs on shirts. Merchants and tailors may find this useful in providing advice to potential customers.3

Discussion

The full description of each shirt is included here in Table 1. To capture a runaway, the most detailed description possible was required. Al­though shirts do not show much variety in color alone, there is considerable variety when other characteristics are taken into account. Oddities such as repairs, ruffles, and buttons may be overstated because of their value in distinguishing the shirt; few descriptions include much detail on these items suggesting that those that are described are unusual. Commonalities in style may be understated. For example, ads [do not] describe tow, normally an unbleached linen, as unbleached; on the other hand, a number ads specifically mention that the tow was white or whitened. Regardless of these nuances, what was common in both color and fabric seems fairly clear since only about 30% of the descriptions did not include at least one of these factors.

When using this data for Caucasian reenactors caution should be employed. Although many of these runaways are apprentices and servants, many are "Negro," "mulatto," "Indian" and "Mustee slaves." In an attempt to get a large population of shirts, ethnic groups were not eliminated from this study. It is left to future authors to separate these populations to see if differences exist.

Not all runaways advertised in Rhode Island newspapers were from Rhode Island. For example, “Barzillae Beckwith, of East-Haddam, in Connecti­cut... had on and took with him... a striped Linen Shirt.” Therefore, these shirt descriptions may be representative of those worn in New England in general, and possibly a wider area.

Some runaways were advertised in more than one newspaper. In these cases, one of the two redundant ads were eliminated but citations are listed for both of the ads. Each of the ads was examined to see if the information was the same, and all available details on the shirts are presented in the table.

Ads for deserted soldiers are not included in the sample. Although many desertion notices appear in the Rhode Island newspapers during the time period in consideration, only seven shirts are described in these ads, too small a sample for useful analysis. Since it is not known whether these shirts were military issue, or were procured privately by the soldiers, they have not been included in this study.

Linen Shirts

In describing runaway’s shirts, fabric was specified slightly more often than color. 71% of ads specify the material (Table 2) whereas 66% of ads specify color (Table 3). Perhaps color was often assumed, especially when fabric was specified.

About a third of the shirts, 34%, were made of linen. There is a great range in the quality of linens, many being coarse unbleached linen of a plain weave, mostly tow (11%) and less often oznabrig (2%).4 An additional three shirts described only as “coarse” might also be tow or oznabrig.

The color of unbleached linens depends on the process used in preparing the linen. The color ranges from dark brown, through tannish, to greyish. No matter what color unbleached linen is, it is never a uniform color but has more of a mottled look.5 Unbleached linens may be bleached by hanging them in the sunlight, as was done with at least some of the tow cloth of shirts described in the ads. Eight shirts were described as “white tow”, as were an additional six in the 1790s. One shirt in 1783 is described as “a whitened Tow-Cloth Shirt” indicat­ing it may not be a true white. Most tow shirts were nonetheless most likely of unbleached linen. That there was a range in coarseness of tow is evidenced by one description in 1777 of “a fine Holland Shirt, a fine Tow Ditto, coarser Ditto.”

Three “striped Tow Shirt[s]” were described in 1767, 1770, and 1782. It is unclear what these unbleached linens were striped with. Possibly they were striped with flax or hemp prepared by two different processes, for example a light gray with a dark brown, but this is only speculation. It is just as likely that these striped tow shirts are striped with white linen. No check tow shirts are described.

Five shirts (2%) were made of tow and linen. This may have been done to make a stripe or check fabric of alternating bleached and unbleached linen. An example of this, in November of 1779, reads, “an Apprentice Lad... about 19 Years old... a Shoe­maker by Trade had on and carried away with him... two Shirts, the one striped Tow and Linen, the other Checks.” Tow and linen could be bleached as with “a Negro Man[s]... white Tow and Linen Shirt” in 1771 and “a Mustee Boy[s]... white Tow and Linen Shirt and Trowsers” in 1784. That this bleaching could occur before the shirt was made is demon­strated by a 1791 ad describing “26 Yards of whit­ened Tow and Linen Cloth” that were stolen.

It is possible that some tow and linen shirts were made with some parts tow and others linen. Al­though no indication of piecing is specifically given for tow and linen it is mentioned for a flannel and tow shirt and a check and white shirt (see the section on Patches/Condition/Oddities). Some shifts are also known to have been pieced with tow and linen.6

11% of linen shirts are described simply as linen. Lacking other descriptors, this is assumed to be a linen of medium coarseness. One linen shirt is described as streaked, two as white, two as checked and five as striped. No hints are given for the colors of these stripes or checks; ads describing striped linens used in other garments show a prevalence of blue and white, followed by dark stripes.

There are a couple of terms used for fine linens including Holland (6%), and “fine linen” (2 shirts). Four shirts described simply as fine are likely of fine linen. Most fine shirts are white as are seven “white Holland” and one “fine white.” Two Hol­land shirts are described as check. The term Irish linen is used only once; since Ireland produced a range of linens from oznabrig to fine linens, we do not know the coarseness or fineness of this shirt.

Wool Shirts

Only a little less common than linen are the 30% of shirts described as woolen. Most of these (24%) are Flannel, a very warm woolen stuff, loosely woven in plain or two-and-two twill. Flan­nel was most commonly white (16 shirts) and striped (14 shirts); it was also checked (8 shirts). The color of the stripes and checks are not given in any of the descriptions except for one “an old black and white checked woollen Shirt” in 1777. (Two flannel shirts in 1795 are described as “striped blue and white,” colors which may be applicable to the 1760-1783 period. A very unusual shirt described in 1786 was “a blue and white Kersey Shirt, with white Sleeve-Gussets, the Neck-Gussets lined with Check” but does not say if the shirt itself is striped or checked.)

Woolen is a generic term for any wool that is carded as opposed to combed. Since flannel is a woolen, some of the 13 shirts described only as “woolen” are may be flannel but could be any of a number of other light woolen fabrics available at the time. Woolen shirts, like flannel, are described as striped (3 shirts) and checked (6 shirts). No color, including white, is given for any woolen shirt.

The only other woolen fabric specifically named is “red baize” for two shirts worn by Negroes, one a blacksmith who ran from Plymouth in March of 1764 the other from “Long-Island, about three months since” April of 1773. These are the only shirts that are described as any color other than white or un­bleached. Baize is a heavy woolen; it is possible that the fabric was actually “bay”, another course woolen more likely to be used for shirts.

Cottons, Blends and Undetermined Fabrics

Only one cotton shirt is described, in September of 1770, but no details are given. It should be noted that cotton is a term used as late as the 1760s, if not later, for purely wool fabrics, and there is some evidence to suggest that linen/cotton blends of British manufacture were sometimes referred to simply as cotton. With no details given about the single shirt described as cotton, we cannot be certain of the type of fabric that it was actually made from.

(A few cotton shirts are described in Rhode Island ads after 1783. One each in 1788 and 1799 give no descriptive details, but an 1800 ad includes “a striped cotton shirt.” From 1730 through 1800, just over 1% of the shirts described in Rhode Island runaway ads are cotton, suggesting that these fabrics were uncommon for the lower sort in New England throughout the century.)

For the remaining 36% of the shirts mentioned in the ads, the fabric is either not described, the description is vague, or it includes blends. Within this group are “cotton and linen” but only three shirts (less than 2%) are described in this way. Two cotton and linen shirts are striped; one carries no other details. The only other blend described is a “checked Flannel and Tow” shirt

Shirt Markings

Marking shirts was usually accomplished by stitching or inking letters and/or numbers below the bosom slit or on the lower part of the front tail. Shirts are often marked with the owner’s initials, sometimes followed by a number. These markings are believed to be used to identify the owner when the shirt is sent to the laundry.

If marking shirts was common practice, it is strange that only one shirt out of 250 mention any marking at all. Several hypothesis for not describ­ing markings are possible. First, the marks may not be known by those writing the descriptions. If the markings are on the front tail of the shirt they would only be seen by the owner and anyone washing the garment. Even if the mark is put below the bosom slit it would not be visible on a man wearing a jacket or waistcoat and neckcloth. Some existing garments in museums are of white linen with white silk marks making it difficult to see them.

Markings can be easily ripped out of a shirt without doing any damage to the shirt. Finally, it is possible that the lower sort had no reason to mark their shirts. Would a poor man pay to have his shirt washed? Would he wash it himself or have a family member do it?

A “Negro Man” whose name is not given ran away from his master “Robert Stanton of Stonington” in 1760. The man “Had on when he went away... a white Flannel Shirt marked C on the Bosom”. The initial certainly is not that of the master and although the Negro’s name is not given, Negroes often went by only one name. Another ad, outside of the time period of this study, describes “several Holland Shirts marked J.S.... belonging to Joseph Simmons.”

Buttons

No buttons on any shirts were described. Most shirt buttons during this time are assumed to be made of white or unbleached linen thread. Some gentlemen did wear what we think of as cufflinks that could be removed during laundering. In two cases “sleeve buttons” are described separately from shirts. Another description includes “one Pair Stone Buttons” that were stolen in 1770. It is possible these “sleeve buttons” or pair of buttons could be worn on other garments but it seems likely they were intended for shirts.

A “Negro Man... about 40 Years of Age: he pretends to be free, is a Clothier by Trade, and sometimes... a Pedler of... Sleeve-buttons” was described in 1772. Another “Negro Man... Had on and took with him when he went away... a Pair of Silver Sleeve Buttons” in 1777.

Ruffles

Often existing shirts from the period have ruffles at the bosom slit and at the wristbands. Most runaway ads did not describe ruffles, but four (fewer than 2%) did. Three simply say “ruffled shirt” and one is described as “fine ruffled.” Ruffles may or may not be noticeable depending upon what other clothing is worn and how it is worn; it is possible that more shirts had ruffles than are mentioned.

Patches/Condition/Oddities

Only one patch is described, “an old striped flannel shirt, patched with linen”. Nine other shirts are described as old. One set of clothes including a shirt is described as “somewhat tarry.” Another shirt in 1775 may have a patch or some sort of repair since it is “a white tow shirt, with a piece put in on one of the sides, up and down.”

Sometimes shirts were pieced together with different fabrics. In 1771 “an indentured Servant Man... an Englishman... Had on when he went away... a Check Shirt, with white Wristbands.” In 1775 “a Negro Man... hath several Marks, left by the King’s Evil, on his Neck, just above the Collar of his Shirt: He took off with him... a white Flannel Shirt, with an unwhitened Tow cloth Collar... Home­spun”. Ready made shirts could be purchased in stores but one shirt in 1782 is described as a “fine Home-made Shirt.”

Conclusions/Interpretations

Although excellent shirt patterns are readily available, the fabrics recommended in these patterns do not include the variability seen among runaways advertised in Rhode Island. It seems that most shirts were white or off white linens as recommended by the shirt patterns. However, unbleached tow and oznabrig were only slightly less common than white in New England.

Bryan Howard wrote what is probably the most comprehensive work on poor men’s clothing to date. In Howard’s Ph.D. dissertation he concludes that among runaway indentured servants advertised in Virginia unbleached brown shirts were probably more common than white. The mounting evidence strongly suggests that unbleached tow and oznabrig shirts with a mottled brown or grayish color were common throughout the American colonies.

Moreover, almost one in every four New Eng­land shirts were wool flannel, either white, checked or striped. Richard Rolt, writing his dictionary in 1761, defines flannel as a “loose, woolen stuff... some use it for... shirts, and shifts”.7 It makes sense that Bryan Howard did not find any flannel shirts in the warm climate of Virginia but reenactors who attend winter events in more northern latitudes will appreciate a soft wool flannel shirt. None of the commercial shirt patterns reviewed in this study recommend wool flannel.

The striped and checked flannels were probably most commonly striped blue and white. Although cotton and linen/cotton blends are recommended by all of the patterns reviewed, both of these fabrics combined account for less than 2% of the shirts described. Likewise only one shirt pattern recommends stripes. This study finds nearly 20% of the shirts were striped, and checks were only slightly less common. It is possible that striped shirts were a regional fashion; Bryan Howard found no stripes among indentured servants in Virginia. A quick search among runaways advertised in Penn­sylvania likewise revealed few striped shirts.8

This study is inconclusive about how common shirt markings were. It is fairly certain that fancy metal buttons were not included on most poor men’s shirts, nor were ruffles. Most buttons were likely to be made of thread, since they did not stand out enough to be described in ads. Some shirts worn by the lower sort were patched with fabrics different from the body of the shirt or pieced together using different fabrics.

Commercially available shirt patterns, in order to be appropriate for a variety of impressions, typically recommend only fabrics that were proba­bly available in any location. These fabrics may not, however, have been common in any region. When using a commercial pattern, it is important to base the choice of fabric on the region and situation being portrayed, rather than simply following the recommendations given in the pattern. Reenactors and interpreters who portray the lower sort in New England should consider a variety of fabrics, includ­ing a mix of old coarse unbleached linen, fine white Holland ruffled, checked wool flannel, coarse whitened towcloth, fine check Holland and striped linen shirts, to name a few.

Last updated: July 26, 2024