In Living Color

Since its opening in 2011, Crystal Bridges Museum of American Art has generated a great deal of interest from the public and other cultural institutions. From the construction of a 185,000-square-foot facility in the bottom of a ravine to the much-discussed art acquisitions, this attention is hardly surprising. However, beyond the building and the art, Crystal Bridges also has an art research library with many rare books and the most important collection of American color printed books in North America. This article focuses on the genres and methods of color printing in the Americas in the nineteenth century, then follows with a discussion of the challenges for this collection, as well as some guidelines for cataloging its items.


american color plate collection
The color plate collection consists of 1,088 items printed in color in the Americas between 1800 and approximately 1910. These items were the personal collection of New Haven rare book scholar, collector, and seller William S. Reese. Reese's interest in color plate books came from his early interest in books of natural history. As he states in an unpublished essay concerning the creation of this collection: My initial interest in American color plate books stemmed from my earliest collecting interest, American natural history. Some of the most beautiful and interesting American color plate books are works of natural history, most famously the smaller-scale octavo edition of J.J. Audubon's The Birds of America and his massive work on animals, the elephant folio Quadrupeds. The former was the first rare book I owned, and it introduced me to some of the potential bibliographical complexities of such a work. In the late 1970s and 1980s, when I had become a rare book dealer specializing in Americana, I continued to add to my natural history holdings with the works of Thomas Say, Alexander Wil-son, Michaux and Nuttall, and W.P.C. Barton. These became the nucleus of the larger collection. 2 Many of the items in the collection are held only by Crystal Bridges, and the collection presents a unique opportunity to compare titles and printing methods side-by-side from a rich array of titles and themes. Reese continues, "After fifteen years of active collecting I had brought together more works than were held by any library in the United States, or the world for that matter-more than the Library of Congress, Harvard or Yale, or any other public or private library." 3 Titles in this collection include some of the most prominent examples of American color printing from the nineteenth century, including several lithographic printings of the work of John James Audubon, the catalog of W.T. Walters's collection of oriental ceramic art printed by Louis Prang, and many lesser-known but quite rare titles. 4 William Reese set general guidelines for the collection: that the books in the collection have four plates printed in the Americas, and that those plates have at least three tones. Furthermore, the books must be printed before trichromatic halftone printing (introduced in 1895) became the dominant method of color printing. However, some examples in the collection are as late as 1910 due to their use of chromolithography. Reese generally excluded two large bodies of color illustrationchildren's books and cartography-as both were major fields of collecting in their own right, with an array of collectors and research in those fields.

genres in color printing
The titles in the color plate collection offer visual examples of the themes of color printing in the United States. Due to the great expense and effort required to print color images in a book during the nineteenth century, these books were not printed simply by happenstance but had a specific purpose to fulfill, be it commercial or scientific in nature. Perhaps the most famous example of this genre is the many editions and titles produced by John James Audubon and his sons. Indeed, color images in a book could present accurate scientific or medical knowledge, as is the case in Alexander Wilson's American Ornithology 5 (discussed below) and G. Spratt's Obstetric Tables, 6 respectively. Obstetric Tables was a copy of the title originally published in London and was unusual in that it depicted female reproductive anatomy with up to five levels of flaps, all printed in color. The text and illustrations were key sources of information for midwives in the nineteenth century. Accurate color reproduction was essential for these fields, and thus the highest quality printers were sought for this type of work.
The federal government was the largest printer of color plates in the nineteenth century. As Reese notes in Stamped with a National Character, the Pacific Railroad Surveys (arguably a "scientific" book) alone required the creation of twenty-one million plates, both color and black-and-white. 7 This highlights one of the primary issues for color printers in the Americas-there were few banks or individuals willing to finance the creation of major color books, unlike in Europe where a large and wealthy leisure class had the capital and interest. The revenue of the federal government was able to finance the largest color printing efforts of the nineteenth century: the Pacific Railroad Surveys, as well as the King, Hayden, Powell, and Wheeler surveys. Indeed, the Hayden survey's chromolithographic reproductions of Thomas Moran's sketches of Yellowstone (printed by Louis Prang) are among the finest examples of American color printing. 8 However, not all work was scientific; many examples in the collection are commercial in nature. Two especially interesting examples stand out: Godey's Lady's Book, and the bound Burpee seed catalogs. Godey's Lady's Book was a women's magazine featuring a wide array of writing, as well as illustrations in color of the latest styles of dress. These plates were generally produced as black-and-white engravings and sent to women in their homes to be colored in by hand, based on instructions, making each plate unique. It can be argued that the Lady's Book was then a rudimentary "by women, for women" publication.
The collection also includes a one-of-a-kind example of commercial publication, W. Atlee Burpee's personal bound copy of his company's seed catalogs for 1901. Bound in limp gilt morocco leather with his name gold-stamped on the front (Figure 1), these three catalogs hold Burpee's notes, suggestions, and price adjustments for the upcoming year's catalogs ( Figure 2). These catalogs include two chromolithographic title pages plus six chromolithographic plates, a process discussed below.
Other themes besides science and commerce are found in the collection. A notable example is the commonplace book of author, lecturer, and editor Emma C. Embury (1806 -1863). 9 Collected for the tipped-in chromolithographic plates, the book also includes original poems by Embury, as well as her pencil sketches and watercolors ( Figure 3) made on her travels. The album also includes illustrations by Anna Embury. Bound in deep purple morocco and gold stamped, this commonplace book is an interesting example with material from a remarkable nineteenth-century American woman.

progression of color printing methods
Beyond the themes of color printing, the progression of color printing methods used in the production of these books and in books throughout the Americas can be seen in the titles in the collection. The methods by which color images were produced in the nineteenth century generally follow an arc from most expensive and laborintensive to least expensive. Parts of the process could be, and were, automated by machinery, but the majority of the work in producing these plates was done by hand. 10 By the turn of the twentieth century, color was common in printed material, largely due to the advent of the trichromatic halftone printing process. The methods used to achieve color in books can be divided roughly into chronological order according to method: engravings, etchings, tinted or colored lithographs, and chromolithography.
As is often the case with any classification, not all titles in the collection fall neatly into these categories. Jacob Bigelow's American Medical Botany 11 is a prime example of this. Bigelow-a physician, botanist, and architect-originally planned to print the plates in his book using copper plate engravings, but he concluded that the process was too expensive. 12 The result is a book printed using a hybrid of printing methods.  The first half of the first volume is printed using copper plate engravings, but the rest of the plates are printed using etched stone inked à la poupée, meaning "with the doll." Bamber Gascoigne describes this term as coming from "the doll-shaped bundle of fabric which is used to dab the ink of differing colours into the grooves and dots recessed in the surface of the plate. In this method the printer is in effect painting the picture on the plate for each impression, and however much care is taken to achieve uniformity each print will be unique in its colouring." 13 engravings The first book printed in color in the United States, William Birch's The City of Philadelphia in the Year 1800, 14 was made using hand-colored copperplate engravings. This process required the engraver to manually remove material from the copper printing plate, creating a surface into which ink could be rubbed. After being printed in black and white, the printed image was then colored by hand or colors were applied to the plate before printing, resulting in a finished full color print from one impres-sion. Printing in this manner was time-consuming, labor intensive, and thus resulted in a far higher price for the book when it was published. Another significant title in the collection printed using this method is Alexander Wilson's American Ornithology, the first American color book to convey scientific information. Wilson set the tone for early scientific and natural history books through the quality and thoroughness of the selection and production of the color plates. Later books, such as Bonaparte's ornithological book, 15 directly referenced and drew upon Wilson's earlier efforts. The use of engraving for color illustrations waned with the advent of etching and lithography,  but some later examples still used this method. Michaux's North American Sylva 16 is a prime example of this, with a captivating history as well. Michaux's study of trees, co-issued with Nuttall's Sylva. . .Not In The Work of Michaux, 17 not only stands as a late example of color printing using engravings, it is also indicative of the fascinating backstory behind the production of some of these books. Originally engraved and published in France, both in 1810 -1830 and in 1817-1819, the latter under a false Philadelphia imprint, the plates were purchased by William Maclure, prominent naturalist and president of the Academy of Natural Sciences in Philadelphia. Maclure brought the plates with him to the utopian settlement of New Harmony, Indiana, where the first American edition of this work was published in 1841. The collection includes this rare printing of the book. Though few of the titles in the collection have such a rich story, they were all the result of remarkable efforts by the authors, publishers, and printers.

etchings
Despite these early successes (the Sylva went into several reprints), printers found the limited ability of engraving to create smooth gradations of color from lighter to darker tones to be less than satisfactory. Furthermore, etching had the potential to be less labor and resource intensive, thus reducing the cost of printing using this method. Engraving was gradually supplanted by etchings, with etchings being the dominant form of color printing in the 1820s. Etchings and engravings both use metal plates for printing, but they each create the negative image on the plate using different methods. The etching uses acid, rather than a tool, to remove material from the printing plate to create the image. Aquatint was far and away the primary method of creating etchings in the United States. An aquatint is created using a metal plate that is covered in rosin dust, then heated so that the rosin melts and creates a uniform surface on top of the metal plate. The area that the printer does not want affected by the acid is stopped out with acid-resistant varnish, and the plate is placed in an acid bath. This acid bath is where the name aquatint originates-the acid used was generally nitric acid, the Latin name for which is aqua fortis. After several baths in acid with various areas stopped out, the resulting image creates a gradual transition from light to dark, as the first areas etched become deeper with each successive immersion in the nitric acid. 18 William Wall's Hudson River Portfolio 19 is arguably the finest example of American aquatinting and is among the finest examples of all American color printing in the nineteenth century. In a uniquely American twist, both Wall and the etcher of the plates, John Hill, were immigrants to the young United States. 20 Their European training in particular contributed to the high quality of the color images in the book. In the course of twenty plates, Wall and Hill take readers from the mouth of the Hudson River as illustrated by "Little Falls at Luzerne" to the river's end, illustrated by "New York, From Governors Island." In this way, the book mirrors panoramas and moving panorama paintings, such as John Banvard's Mississippi River panoramas and Peter Grain's Panorama of the Hudson and James Rivers-Scenes in Virginia. A less serious, though still notable, work printed using aquatint was the pirated 1819 Philadelphia printing of The Tour of Dr. Syntax in Search of the Picturesque. 21 The English satire featuring Dr. Syntax in many ludicrous mishaps was popular in the United States, so much so that the printer found it profitable to pirate and print the second volume in the Syntax series in 1822. 22

lithography in general
By the end of the 1820s, lithography was rapidly becoming the most widely used method for creating color images in books. This popularity stems largely from the adaptability of lithographic printing to some form of automation and mechanization. Indeed, books printed using lithography represent the bulk of titles in the color plate collection. To create a lithograph, an artist prepares the surface typically using what can be described as a greasy pencil to block out areas to repel ink. After applying the pencil, the stone is "fixed" and then inked and printed. This fundamental process went through many iterations in the United States, from hand-colored lithographs, to tinted lithographs, and finally to chromolithography.
The three main types of lithography practiced in the United States all shared printing from stones as described above, but differed in the manner in which color was added to the initial black-and-white image. A hand-colored lithograph, such as those in Doughty's Cabinet of Natural History, 23 would be printed in black and white and colored by hand after the fact, much like hand-colored engravings. A tinted lithograph began in the same way with a black-and-white image on top of which would be added tints from other lithographic printing stones. Multiple tints could be used to provide several colors in one image. The advent of high quality chromolithography in the 1850s created images that were comparable in quality to hand-colored plates at less cost. Chromolithography typically used twelve plates to create a finished color plate. Black would be the first color, but after that, sequencing the colors, choice of specific colors, and registration of the paper on the plate were the purview of craftspeople called chromistes. Even at this late stage with less manual work needed, there was still a significant requirement for professionals trained in the sequencing, coloring, and registration of the lithographic stones. According to Suzanne Low, with just a glance at a color plate, the chromistes could tell how many and which colors were used in the creation of a finished color image. 24 (Figure 4) is a breathtaking work to see. Walters compared Prang's work to that from other prominent color printing firms, including those in Europe. Walters decided that Prang produced the highest-quality prints, and Prang was thus commissioned to create the full catalog in a limited edition of 500 copies for ten volumes. While most chromolithographic printers used twelve stones for a completed image, in order to meet Walters' requirement for color images of the highest quality, the plates in this book used from twenty to forty-four separate stones for each image. 26 Perhaps understanding that this work was the apex of this type of printing and that new methods for printing color were coming into common use in the United States, Prang merged his business with the Taber Art Company in 1897 and retired. 27

chromolithographic nursery catalogs
While not as visually impressive as the other books, the William S. Reese collection also includes a significant number of nursery salesman's catalogs printed in Rochester, New York. Reese describes these catalogs in this way: Another unusual sideshow in color plate illustration was the printing industry that developed in Rochester, New York, in the 1850s. The booming nursery business of that city needed color plates to illustrate the fruits and plants for sale there. The first plates, produced by stencil and hand coloring by firms such as Sargent, are perhaps closer to folk art than color printing. These plates were seldom issued as books in editions, but were usually gathered up to suit the individual interests of a customer. As a result, almost all of the Rochester catalogues are unique in their collations. Stencil and hand coloring was eclipsed by chromolithography by the 1870s. 28 These nursery catalogs are illustrative of the larger cataloging challenges for this collection. Tracing the different printing firms, as well as each printing process used, was important for facilitating access to the collection through the library's online catalog. In addition, some bindings were quite notable, and some items had important associations with prominent persons, such as the Burpee catalogs described above. When the museum acquired this collection, Reese also included a database with his notes for each of the titles, providing expert background, commentary, and descriptions for some of these rare and unique titles. Several pertinent points in Reese's text above provide important information for the cataloger. First, the plates themselves were printed by a variety of firms in a wide variety of processes. Second, the uniqueness of each item the cataloger has in hand has much to do with the needs of the company collating the catalog, as well as the variety in processes from the different printing firms in Rochester 29 ( Figure 5). Third, there was no standard process to create the color plates, and indeed, they can vary within the item itself from hand-colored images, to photolithographs, to hand-colored photographs. Finally, the plates were printed in Rochester, New York, and assembled by firms in a variety of places, depending upon the company's needs. The unique collations and color in these items are largely a result of their purpose: to equip door-to-door nursery salesmen with the appropriate items to sell their goods.
Color printing in the United States at this time was perhaps more common than at the beginning of the nineteenth century, but it was still relatively rare compared to books illustrated with monochromatic prints. Color was necessary to accurately depict the plants and fruits for sale, and the rarity of color was likely an added inducement to purchase. Indeed, the expense of these catalogs is often implied through the admonition to the salesman to take great care with the book, lest it be removed from the salesman's pay. The flyleaf of one of these catalogs states: "Please handle Plate Book carefully, and do not mark prices on the plates. Keep it out of the hands of children, and avoid getting it wet or soiled, as the books are expensive, and must be protected from all unnecessary damage." The often intricate flaps and bindings show the additional cost borne by some nurseries to protect the plates within to the highest degree possible. Indeed, these efforts helped to ensure that some examples survive today, even after hard use traveling door to door.

cataloging recommendations for nursery catalogs
What follows is a brief outline of guidelines drafted by the author to ensure consistent and accurate description of these items at the Crystal Bridges Museum library.
• Title ⅙ The items in this collection all lacked a chief source of information, or a statement of responsibility, which is in keeping with the nature of these catalogs. Therefore, all the items in this collection have a supplied title, as instructed in AACR2 30 rule 1.1B7: y Supply a title proper for an item lacking a chief source of information from the rest of the item, or a reference source, or elsewhere. If no title can be found in any source, devise a brief descriptive title. Enclose such a supplied or devised title in square brackets. ⅙ The discrete collection of which these catalogs are a part was supplied with a database of bibliographic information about the items in that collection. The title information is taken from this database. ⅙ Therefore, we have a 245 field, as main entry and bracketed.
• Publisher Information ⅙ Publisher information, much like the title for these items, must be deduced from the item, as well as from the accompanying database. The publisher for these items is established in one of three ways. y If the catalog itself has the name of a nursery on it, the name of the nursery itself is taken as the publisher. This decision was made as the nursery was responsible for the assembly of the content, etc.-taking the role of a publisher. y If no nursery name is found on the item, then the publisher information is taken from the printer identified on the plates. y If neither of the above applies, then AACR2 rules 1.4C6 and 1.4D6 are followed, resulting in the use of the abbreviations S.n. and s.l. y In any case, the date of publication was taken from the database information supplied with the collection, and the information was bracketed, as none of the information came from a recognized source of information in the item.
• Physical Description ⅙ All of these catalogs lacked pagination, and in normal circumstances, LCRI 2.5B7 would be followed. However, owing to the rare and unique nature of these items, a bracketed count of pages was inserted, as instructed in AACR2 rule 2.5B7: y If the pages or leaves of a volume are unnumbered and the number of pages or leaves is readily ascertainable, give the number in square brackets. If the number is not readily ascertainable, estimate the number of pages or leaves and give that estimated number without square brackets and preceded by ca. ⅙ Illustrative matter consisted solely of the color plates, so "col. ill." or "ill.
• Notes Fields ⅙ Due to the variance in the methods used to produce the color illustrations, a note was added by the cataloger to give additional detail about the process used to create the color plates, as well as (in some cases) the firm responsible for the printing of those plates. This information provides a basis for the added entries and genre terms in the record, and also gives some enumeration to the illustrative matter. ⅙ On occasion, a note about the binding is added in a 563 MARC field if the binding is unusual or notable. Some catalogs (as discussed in the background section) feature a clasp, or gold and blind stamping. This information is included in the field, along with the appropriate institutional code in subfield 5.
• Subject Access ⅙ Subject access and terms were generally consistent in all records for the nursery catalogs. The terms varied depending upon what was illustrated in the catalog (trees, fruit, flowers, etc.) but the final subject heading Nurseries (Horticulture)-Catalogs was always present. Also, the free floating subdivision Catalogs appears after each term, as a "c" is present in the "Cont" fixed field.
• Genre Term Access ⅙ In the appropriate MARC field, printing terms are traced using appropriate thesauri. Typically, terms from the Library of Congress's Thesaurus for Graphic Materials 31 were used, but other terms were used from the Rare Books and Manuscripts thesauri. 32 Furthermore, when appropriate, the binding and other related terms were described in this area.
• Added Entries ⅙ Entries were traced for these categories: y Lithographers/printers. Entries tracing the firm responsible for the printing of the plates were always entered. y Nurseries. On occasion, if they were not traced as the publisher, the name of the nursery itself was traced in this area. y Illustrators. Finally, if the illustrators could be determined, they were traced in the 7XX field as well. ⅙ Added entries have either relator terms or relator codes to describe the relationship of the entry to the item being cataloged.

conclusion
A yeoman's effort was required of the two librarians at Crystal Bridges to bring the library into existence in just over a year. From the unpacking and integration of 50,000 items, to working with a third party vendor to have 30,000 titles ready for opening day, to the more minute tasks associated with cataloging and metadata, much work was undertaken and knowledge gained. The resulting library, accessible and open to the public, has had over 13,000 visitors, some enjoying the view of the museum and surrounding area and others undertaking scholarly research in a variety of disciplines. The library's unique collection of color books printed in the Americas in the nineteenth century is an unmatched resource for researchers, scholars, and individuals interested in color printing. The color plate collection in particular has fascinating connections throughout the history of the United States as well as a rich array of primary and secondary source materials. The titles in the color plate collection are available by appointment any time the museum is open.

acknowledgments
This article would not have been possible without the learned assistance of Ron Tyler, former director of the Amon Carter Museum of American Art. Also essential was cataloging advice and mentoring given by Penny Baker, collections management librarian at the Sterling and Francine Clark Art Institute; Samuel Duncan, library director, and Mary Jane Harbison, library technical services coordinator, both at the Amon Carter; and the inimitable Sherman Clarke. Finally, this article would not have come to fruition without the encouragement and assistance of Jennifer S. Dean. The author owes them all a deep debt of gratitude.