Edwards issued 152 splendid color plates of North American butterflies, but he failed to comprehend the difficulties that he would encounter along the way. The project dominated over thirty years of his life and required the sale of his beloved butterfly collection to generate the necessary funds. The production of the book was tremendously expensive to Edwards, both financially and emotionally. It came to symbolize the end of an era, when fine hand-colored illustration was being replaced by more efficient and inexpensive methods of color lithography and photography. As early as the 1920s, the third volume was described as “ excessively scarce,” while complete sets were considered quite rare (Sherman 1925). Today, the book seldom surfaces in the antiquarian book market. When it does, copies are expensive and usually incomplete. Rare complete sets of the book now sell for $7,000–$10,000 US. Despite the book's significance, I was surprised to learn how little is known about its history. Published information is scanty and largely inaccurate. I recently renewed my fascination with this influential work and decided to conduct a long-overdue investigation into its production.

Fig. 1.

William H. Edwards, ca. 1880, during the production of the second volume of BNA (from Bethune 1909).

This is the story of the creation of The Butterflies of North America, gleaned mostly from Edwards' writings and conveyed as much as possible in his own words. To call Edwards a prolific writer is an understatement; thousands of his letters and postcards are preserved in various museums and libraries. He wrote so many letters that he often repeated information to the same correspondents without realizing his redundancy. Edwards' inexhaustible pen documented a great deal of the book's production. This study does not attempt to explore the numerous taxonomic quandaries or other aspects of Edwards' entomological labors. Useful summaries of Edwards' life and contributions were offered by Bethune (1909), dos Passos (1951), Mallis (1971), Sorensen (1995), and Weeks (1911). Leach (2013) considered Edwards within the broad context of 19th century natural history and examined how his discoveries significantly impacted the study of Lepidoptera in America.

Methods

Multiple copies of The Butterflies of North America (hereafter referred to as BNA) were analyzed, including my own. Also consulted were several thousand unpublished documents, mostly from Edwards' correspondence. Photocopies and scans of letters, postcards, and other manuscripts were obtained from the following: Research Library of the American Museum of Natural History (New York, New York: AMNH), Library of the California Academy of Sciences (San Francisco, California; CAS), Ernst Mayr Library (Museum of Comparative Zoology, Harvard University, Cambridge, Massachusetts; MCZ), Lyman Library (Museum of Science, Boston, Massachusetts: BMS), McGuire Center for Lepidoptera and Biodiversity (Florida Museum of Natural History, Gainesville; MGCL), Ewell Sale Stewart Library (The Academy of Natural Sciences, Philadelphia, Pennsylvania; ANSP), Archives and Special Collections, Olin Library, Rollins College (Winter Park, Florida; RC), Charles C. Wise, Jr. Library (West Virginia State Archives, West Virginia University, Morgantown; WVSA), and the personal library of James R. Wiker (Greenview, Illinois).

Extremely valuable were photocopies, microfilm prints, and photographs of hundreds of letters and postcards which were collected from many sources during the 1960s by the late F. Martin Brown. Initially donated to the Allyn Museum of Entomology (Sarasota, Florida), these copies are now preserved in the archives of MGCL. The original manuscripts are deposited in the AMNH, MCZ, and WVSA, as well as the library of the Carnegie Museum of Natural History (Pittsburgh, Pennsylvania; CMNH), the library of the Field Museum of Natural History (Chicago, Illinois; FMNH), and the Smithsonian Institution Archives (National Museum of Natural History, Washington, D.C.; USNM).

Among the manuscripts studied were letters and postcards from W. H. Edwards to Spencer F. Baird (SB) (1824–1887) (many of these were published by Brown (1958, 1959a, 1959b, 1960)), Henry Edwards (HE) (1827–1891), Hermann A. Hagen (HH) (1817–1893), William J. Holland (WH) (1848–1932), Edith K. A. Mead (EM) (1852–1927) (W. H. Edwards' daughter), Theodore L. Mead (TM) (1852–1936), Carl R. Osten Sacken (CS) (1828–1906), Samuel H. Scudder (SS) (1837–1911), Henry Skinner (HS) (1861–1926), F. H. Herman Strecker (FS) (1836–1901), Philip R. Uhler (PU) (1835–1913) and William G. Wright (WW) (1830–1912). Letters and postcards to W. H. Edwards include those from Louis Agassiz (LA) (1807–1873), Henry W. Bates (HB) (1825–1892), Charles J. S. Bethune (CB) (1838–1932), Augustus R. Grote (AG) (1841–1903), John Hamilton (JH) (1827–1897), Gilbert M. Levette (GL) (1833–1889), and John W. Weidemeyer (JW) (1819–1896). Also examined were 49 original letters from W. H. Edwards to T. L. Mead which were recently discovered in the archives of the McGuire Center. These letters were obtained in 1981 by F. M. Brown from the great-grandson of Edwards.

Because of the enormous amount of correspondence consulted, I have chosen to not cite all the letters and postcards parenthetically within the text. Instead, I cite each with a superscript number which corresponds to an entry in Table 1. For brevity, each letter citation is formatted using the correspondent's initials (as given above), followed by the recorded date (day, month, year), and the acronym of the repository of the original manuscript (as above). Each entry is prefixed by a lower case letter to indicate whether the correspondence was sent to (“t”) or from (“f”) W. H. Edwards. For example, “tWH 12.xii.1885 CMNH” refers to a letter sent to W. J. Holland, dated 12 December 1885, deposited in the Carnegie Museum of Natural History. The citation “fSS 7.ix.1882 WVSA” refers to a letter sent from Samuel H. Scudder, dated 7 September 1882, deposited in the West Virginia State Archives.

Table 1.

Cited correspondence of William H. Edwards by section.

continued

continued

A photocopy of W. H. Edwards' typescript autobiography (Edwards 1900–1901, 1902) was generously furnished by Douglas M. Willis, great-greatgrandson of Edwards. It includes much information not given in Edwards' “Entomological Reminiscences,” published by dos Passos (1951).

In an attempt to determine dates of issue for the parts of BNA, I searched numerous publications for advertisements, announcements, library donation lists, and reviews that alluded to specific parts of BNA. State and federal census records, city directories, and burial records were consulted to reveal information about the various people involved in the creation of the book.

Results

Conception and tragedy. About the year 1865, William Henry Edwards resolved to publish a series of colored figures of North American butterflies of the genus Argynnis (=Speyeria). He announced, “I propose publishing a monograph of the Argynnides of this country after the style of Hewitson's Exotics” [i.e. early parts of Hewitson 1852–1877].¹ Edwards soon decided to alter his original concept and include additional genera, publishing what he could and letting “the numbers run along” until he made a full volume.² His initial plan was summarized in a “Prospectus,” in which he proposed to issue a book in installments (parts) at three-month intervals and offer it for sale at $2.50 per part. Each part would contain several hand-colored plates, with accompanying letterpress, which portrayed species “mostly new, or, if old, those that have heretofore been incorrectly described or figured” (see Edwards 1869). Edwards initially proposed that his book be titled, “Illustrations of the Butterflies of the U.S. & British America,” but S. F. Baird of the Smithsonian Institution suggested other alternatives, which Edwards adapted into the final title, The Butterflies of North America: With Colored Drawings and Descriptions.

Edwards struggled with the content of the text, asking Baird if he should “say anything of the habits/localities” or confine himself strictly to descriptions.³ Edwards ultimately settled on a format that incorporated all this information. He considered segments of text to represent “papers,” similar to what would be published in a scientific journal.⁴ He also decided to issue a “ Synopsis of Species” in installments, which would be concluded within the volume (Edwards 1869). This Synopsis of North American Butterflies was furnished separately as the sheets appeared for $1.00 extra. The first part of BNA was published “as an experiment” in 1868 and its success encouraged Edwards to continue. During the summer of 1868 he proudly declared, “My book is taking well I believe, & appears to be appreciated.”⁵

Edwards began writing the text for BNA at his home in Newburgh, New York. In late 1868 he permanently moved to Coalburg (formerly “Coalburgh”), West Virginia, to more effectively manage the Kanawha and Ohio Coal Company, which he co-founded several years before (Edwards 1902). For two years he and his family lived in a small house, after which they moved into a larger two-story home on another property. Edwards, however, was frustrated about his relative isolation in West Virginia. “I work under great disadvantages in the absence from Libraries and from the artists and printers employed,” he wrote. “In this remote corner of Virginia the mails are often ten days from New York or Phila.”⁶ West Virginia had become a state only seven years earlier, having seceded from Virginia during the Civil War. Edwards and his family often wintered in Charleston, West Virginia, located about 22 km (13.7 mi) northwest of Coalburg.⁷

On 21 February 1871, during the production of the first volume of BNA, Edwards' home was consumed by fire. Edwards recorded this traumatic incident in his autobiography (Edwards 1900–1901):

Edwards attributed the fire to “the pipe of our colored boy, Wesley Bowles, who was in the habit of smoking after his dinner” (Edwards 1902). The home was a total loss and Edwards carried no insurance.⁸ Edwards' butterfly collection was saved, but was “more or less injured” because the small boxes in which he kept his insects were “grabbed by all sorts of fellows & deposited on the grass at a distance.”⁹ Edwards went in search of his collection “with anxious heart,” and managed to find nearly everything, which actually suffered very little damage. Relieved, he supposed it may have been irreparable “if the fire had broken out in the night & my collections pitched out in the darkness.” Edwards' collection in West Virginia contained his rarer species, as well as all those intended to be used for figures for BNA “for some time to come.”¹⁰ Its loss certainly would have terminated production of BNA. Edwards kept his “old collection” in New York, and did so for many years.¹¹ A brief notice about the destruction of Edwards' home was published by Bethune (1871).

Edwards quickly rebuilt his home, which was “on a different plan from the other & a prettier one.”¹² Edwards' granddaughter, Catherine Tappan Willis (nee Smith) (1884–1968) lived in the home for much of her life and spent a great deal of time with Edwards. She vividly described the home's immediate surroundings: “ Mr. Edwards built his house at the mouth of a wooded ravine where the land sloped down to the Kanawha River. The first settlers had cleared away several acres, leaving a few large trees, but the forest covered the mountain down to the clearing. The place had a natural setting of mountain, hillside and river. … They laid out their garden carefully with beds and borders full of old fashioned flowers, roses and gay annuals to attract butterflies. While fruit trees, berry bushes, and vegetable garden were planted on hillside and river bottoms” (Willis 1901). Thus was the setting in which Edwards conducted his entomological research and produced the three volumes of BNA. A black and white photograph of the home, with Catherine standing on the porch, was published by dos Passos (1949). I obtained a more recent image (Fig. 2) from Leigh Mann, Catherine's great-granddaughter. Through the dedicated efforts of Catherine's son, John A. Willis, Jr., and his wife, Harriet, the house was placed on the National Register of Historic Places (NPS 1990). The property has served as home for five generations of the Edwards family. Today, Edwards' great-great-grandson, Thomas O. Willis (great-grandson of Edwards' youngest daughter, Anne), resides in the house, which sits on a property roughly 3.2 ha (8 ac) in size (NPS 1990). This represents only a tiny fraction of the 34,641 ha (85,600 ac) of land that Edwards once owned in the region, much of which he sold during the 1850s for a handsome profit (Edwards 1900–1901, 1902). Regrettably, strip mining for coal, the unforeseen progeny of Edwards' own local mining operations, has removed the summit of the mountain directly above the house (T. O. Willis, pers. comm.).

Contemporary works. BNA is often compared with another outstanding 19th century book, The Butterflies of the Eastern United States and Canada by Samuel H. Scudder (1888–1889). Scudder was an erudite systematist, while Edwards described a greater quantity of taxa and documented the life histories of countless butterflies. Edwards strongly disagreed with Scudder's use of generic names that were derived from the works of the early German entomologist Jacob Hübner. Scudder's insistence on these names fueled many heated discussions with Edwards and others within the entomological community.

Although Scudder's opinions often diverged from those of Edwards, he published anonymous reviews of BNA at least eight times (Mayor 1919) and his meticulous critiques reflected the overwhelmingly positive reception of the book. Although Scudder was a literary and scientific competitor of Edwards, these entomologists enjoyed a fruitful correspondence. Scudder visited Edwards at his home in West Virginia and even helped obtain funding for the production of BNA. They often sought advice from one another and freely shared specimens of insects and plants, as well as proofs of illustrations for their respective books. Edwards loaned Scudder many original drawings and allowed him to reproduce some figures from BNA for Scudder (1888–1889, 1895). Edwards also published advertisements to Scudder's book, which were printed on the rear wrapper of at least two parts of BNA. Despite their disagreements, these zealous naturalists shared a lengthy relationship dominated by mutual respect (Leach 2013).

Edwards was not on such friendly terms with another entomologist who was issuing a book concurrently with BNA. Between 1872 and 1878, Herman Strecker published 15 parts of his Lepidoptera, Rhopaloceres and Heteroceres, Indigenous and Exotic. Strecker, who was known for his acerbic rhetoric and discontent with other entomologists, attempted to undermine Edwards by being the first to describe and figure new species. This dispute was familiar to their shared correspondents. The lepidopterist Henry Edwards advised Strecker, “I do not think it is good either for yourself or for science as Edwards' book was first in the field, & he is likely to get the earliest and most abundant help” (11.x.73 FMNH). Strecker nonetheless continued to seek specimens from the same sources as W. H. Edwards. The quest to acquire and name new species was highly competitive. Squabbles were frequent and often spilled over into the published literature.

Fig. 2.

William H. Edwards' former home in Coalburg, West Virginia, ca. 1991 (courtesy Leigh Mann).

Publishers. It was suggested by S. F. Baird that Edwards employ a high profile publisher for BNA, such as B. Westermann & Company or Hurd & Houghton, both of New York.¹ Edwards instead decided to publish it through the Entomological Society of Philadelphia,2 which changed its name in 1867 to the American Entomological Society. Subscriptions for the book were directed to Ezra T. Cresson, Sr., who was one of the founders of the Entomological Society of Philadelphia. The text was printed by Cresson and another member, Charles A. Blake, on a hand press located in the society's room at The Academy of Natural Sciences, Philadelphia (Edwards 1900–1901; dos Passos 1951). Cresson rolled on the ink, while Blake, being the stronger of the two, pulled the press (Anonymous 1903). After the first volume was completed, Cresson was no longer connected with the publication of BNA. Cresson was “ obliged from the state of his health” to stop doing work for the American Entomological Society, forcing Edwards to find a new publisher.³ Edwards therefore made arrangements with the firm now known as Houghton, Mifflin and Company to publish a second volume. Edwards paid the cost of publishing the book, while Houghton Mifflin covered all costs associated with production and marketing, deducting a ten percent commission for their efforts.⁴ Needless to say, BNA was a very expensive affair for Edwards. Although Cresson no longer acted as the publisher, he continued to accept orders for the second volume, at least for the early parts.

During the production of the second volume of BNA, contemporary advertisements for Parts 1–4 directed subscribers to either the Boston firm of Houghton Oscar Houghton & Company, or the New York firm of Hurd & Houghton. At that time, the latter was considered the publisher, while the former controlled the manufacture of books (RP 1899). The title pages of these parts identified the publisher as Hurd & Houghton. In 1878, Houghton Oscar Houghton became Houghton, Osgood & Company, and the name Hurd and Houghton was discontinued. As a result, Parts 7 and 8 of BNA bore the name of Houghton, Osgood & Company. The company name changed again in 1880 to Houghton, Mifflin & Company, thus this name was used on the title pages of Parts 9–13. All parts of the third volume carried the publisher's imprint of “Boston and New York: Houghton, Mifflin & Company”. The second and third volumes were printed at Riverside Press (Cambridge, Massachusetts), which was established by Henry Houghton as a production facility for his book business. The individual at Riverside Press who arranged printing of the third volume was apparently named Weise.⁵ Beginning with Part 2 of the second volume and continuing throughout the production of the third volume, publisher's imprints also referred to the London publisher Trübner & Company, who were American literary agents located at 60 Paternoster Row in London. Trübner's sold BNA from the very beginning, even when it was published by the American Entomological Society (Anonymous 1868). Consistent with today's marketing practices, advertising of the book increased prior to the holiday season.

Early parts of the first volume were offered at a price of $2.00, but increasing production costs required that beginning with Part 3 new subscribers paid $2.50. This price is equivalent to about $40 today. Depending upon the number of plates included, parts of the second volume were available for $1.50–$3.50 each. Early parts of BNA were offered by Trübner & Company in London for 10 shillings each (Anonymous 1869), while later parts sold for 12 shillings (TC 1889).

Design and layout. Each part of BNA was issued in sewn signatures with paper wrappers (covers) that displayed the purported date of publication. The parts were mailed flat between pasteboards to prevent damage. The entire title as printed on the wrappers was The Butterflies of North America: With Colored Drawings and Descriptions. The unnumbered pages and plates were printed in quarto size (approx. 23 × 28.58 cm/9 × 11.25 in), dimensions which were comparable to other Lepidoptera books that Edwards had examined.¹ Incorporated into the last installment of each volume were a title page, author's preface, and a “ Systematic Index” which enabled the binder to arrange the plates into their proper order. A “Dates of Issue” sheet also was included, which listed all the parts for that volume and the dates when they were purportedly published (Scudder had suggested the inclusion of this page for each volume2). A “General Index” was provided at the completion of the third (last) volume, which indexed the entire work.

The text was printed using the stereotype method, a decision that Edwards ultimately regretted. More expensive than moveable type, stereotype allows the printer to produce small runs over great periods of time without having to reset the type each time. However, because each letter was not a single piece of type, changes could not easily be made without re-casting the entire page.

Upon completion of each volume, it was recommended that the plates and pages be numbered in pencil according to an “Alphabetical Index”, which was also issued with the last part. This elicited complaints from subscribers (e.g. [Scudder] 1874) who thought the lack of numbering was awkward and cumbersome. Edwards defended this format, telling H. A. Hagen in 1882, “If you will look at Hewitson's Ex. But. [Hewitson 1852–1877] you will see that I followed his example and at the end of his introduction to Volume 1 he says he publishes an index for the binder, and also an Alphabetical index of which the volumes may be numbered on pages with pencil. I follow him in all this.”³ Edwards insisted, “I don't see any better or any other way than the one I adopted.”⁴ This format allowed Edwards the flexibility of issuing plates in whatever order he chose, without regard for taxonomic consistency. It was left up to the subscribers to bind the parts into proper order when the volumes were completed. Although this strategy benefited Edwards, it remained contentious among subscribers for many years. As a concession, Scudder suggested that a general index also be included, which Edwards accepted with some reluctance.⁵

Subscribers typically bound each volume when they received all the parts. Late subscribers often purchased the first two volumes complete, and then bound the parts of the third volume to match. The original wrappers were usually discarded prior to binding, thus very few copies of the book preserve them. Rarer still are several publisher's notices which were issued separately or were printed on the rear covers of the wrappers.

Production. Edwards endeavored to make the “ artistic execution” of the plates for BNA as perfect as possible, “sparing no expense or trouble for this purpose.”¹ To this end, he decided to take advantage of stone lithography, which imparted far greater detail than customary engraving methods. He wrestled, however, with the actual production of the plates and found the greatest difficulty “in getting the figures drawn” on stone.² Many years later, Edwards wrote about the early development of his first plates (Edwards 1900–1901; dos Passos 1951), but his recollections were not entirely consistent with his correspondence, which documented the events as they unfolded. The following is primarily extrapolated from Edwards' letters to S. F. Baird.

Edwards first approached John Cassin (1813–1869) about “issuing a trial plate or two,” thereby “leaving it to the future to determine whether a general work on butterflies would be advisable” (Edwards 1900–1901). Cassin was a talented lithographer and co-proprietor of the lithographic firm of Bowen & Company in Philadelphia. He was also a renowned ornithologist, who served as the Curator and Vice-President of The Academy of Natural Sciences of Philadelphia (Stone 1901).

Around April 1865, the first two trial plates for BNA were drawn on stone by “an Englishman” under the employ of Cassin (Edwards 1900–1901). One portrayed the female of Speyeria diana (Cramer) and the other the male and female of Speyeria atlantis (W. H. Edwards). Edwards obtained a third plate by February 1866.³ Based on the numbering system subsequently employed for these plates, the third illustration possibly portrayed Speyeria cybele (F.). Edwards hoped to have two more plates drawn before issuing the first part of BNA.⁴ Two months later, Edwards received a proof of a fourth plate,⁵ ostensibly depicting Speyeria aphrodite (F.). However, Edwards observed that two of these four plates were “imperfect” and needed to be redone.⁶ Subsequent events suggest that he was referring to the plates of S. diana and S. atlantis. To color the early proof plates, Edwards evidently used Susan Clark Gray (1821-?), wife of the artist Henry P. Gray of New York.7 Edwards met her in 1857 when she boarded at the same Massachusetts hotel (Edwards 1900–1901).

To help alleviate Edwards' anxiety about delays in preparing his plates, S. F. Baird suggested that Edwards try John H. Richard (1807–1881), an artist and colorist from Philadelphia who had worked at the Smithsonian.8 Edwards also briefly contemplated using the Londonbased entomological artist Edward W. Robinson (1835–1877), who had previously worked on illustrations of American sphinx moths, a project begun in 1862 with Edwards' involvement. These moth plates were later published in a limited edition by Weidemeyer et al. (1903). However, Edwards had misgivings about sending specimens overseas, writing, “ I should be sorry to send those fine things over there and have them come back ruined.”⁹ In the end, Edwards did not use either of these artists for BNA.

The two plates created in 1866 (S. cybele and S. aphrodite) were drawn by Daniel Wiest (1842–1901), a Philadelphia artist who was introduced to Edwards by E. T. Cresson, Sr. (dos Passos 1951). Little is known about Wiest, but city directories and census records reveal that he was born in Bayern, Germany and lived with his wife and daughter at 1222 Stiles Street in Philadelphia. His occupation was variably listed as artist, lithographer, and engraver. He is likely the same Daniel Wiest who was elected in 1861 as a member of the Entomological Society of Philadelphia (Anonymous 1861). Wiest is best known for creating the popular lithographic illustration “In Memory of Abraham Lincoln: The Reward of the Just”, which was issued in 1865 by the Philadelphia publisher William Smith. Like E. W. Robinson, Wiest had previously rendered plates of sphinx moths, published decades later by Weidemeyer et al. (1903). He later drew plates for several other publications about Lepidoptera, including Robinson (1869). Wiest's plates for BNA were struck (printed) by Bowen & Company (Edwards 1900–1901).

In early 1867, Wiest was asked to render a new plate of S. diana, this time portraying both the male and female.¹⁰ Edwards described the results as “ handsome.”¹¹ A month later, Edwards' remarked, “I find that the artist who drew my 3 plates of Argynnis [presumably S. cybele, S. aphrodite, and the revised S. diana], is willing to go ahead, and I will therefore have 2 more drawn.”¹² Wiest ultimately completed only one additional plate, that of Speyeria nokomis (W. H. Edwards).

The completion of Wiest's four plates took over a year and Edwards grew tired of waiting. He had hoped to complete all the plates and issue the first part of the book during 1867.¹³ Edwards therefore decided to utilize the earlier plate of S. atlantis, combining it with the four created by Wiest, thus reaching his goal of five plates for the first part. The style of the plate of S. atlantis is more primitive than those rendered by Wiest and was not signed. “I believe I have 5 plates of the Agynnides now drawn ready for coloring,” Edwards wrote, “and I have nearly prepared the text to accompany them.”14

As the first part of the book was nearing completion in early 1868, Edwards revealed, “I have made full arrangements for the issue of No. 1 of the ‘Butterflies of N. Ama about 1st Ap[ril].”¹⁵ Still running behind schedule, Edwards finally issued the five plates, with accompanying letterpress, during the first week of June. Using the genus recognized at that time, the plates were entitled Argynnis I - Argynnis V. Looking forward to a second part, Edwards wrote, “If I can get Weist [sic] to proceed with drawing I will put in his hands 5 more.”¹⁶

Although the first part of BNA was well received, Edwards was still disappointed with the quality of his plates. Rather than continue with Wiest as he had proposed, Cassin suggested another artist employed by Bowen & Company, the very talented Mary Peart (1837–1917), who resided about 2.4 km (1.5 mi) from The Academy of Natural Sciences in Philadelphia. Peart was engaged to draw plate Argynnis VI (Speyeria callippe (Boisduval)), which was much more satisfactory. Peart was thereafter engaged to draw all the remaining plates for the first volume, as well as all those for the second. Describing Peart's early work, Edwards (1900–1901) remarked, “She, at the beginning, knew nothing of butterflies, but made drawings of the material put before her. I called her attention to the peculiarities of the legs and antennae and gave her a net with which to take live butterflies in order to study these organs, and soon her drawings became exact.” Edwards later considered his discovery of Peart “as important a find as a new planet almost.”¹⁷ For many years thereafter Edwards declined employing any other hand to draw his plates, even though it delayed production of his book.18

Edwards sent Mary Peart adult butterflies, as well as specimens and sketches of early stages. To reduce the possibility of damaging adult specimens during the preparation of the plates, Edwards sometimes fastened them into special boxes with two glass sides, “in which the insect is placed on cork.”^19,20 He described these boxes as “6×8, 2 inches deep inside . . . one half shuts down on the other, tight.”²¹ Such boxes, similar in construction to those Edwards used to store portions of his own collection, were apparently inspired by the storage boxes of the Philadelphia naturalist Titian R. Peale, whom Edwards met during the 1860s (Edwards 1900–1901). The specimens thus protected, Peart was “ able to work just as well so as if she had the naked insect to handle.”²²

While the purpose of the first volume of BNA was intended to illustrate species that were “incorrectly described or figured,” Edwards' attention ultimately turned to documenting the early stages of each species. Although this added a great deal of complexity to the project, it significantly improved the usefulness of the book. Edwards urged others to send drawings and specimens of early stages to Peart for illustration, sometimes offering to exchange copies of his book for their help: “[I]f that will not fetch it, I don't know what will.”²³ Edwards could be quite insistent in his efforts to encourage correspondents to provide material, even scolding Henry Edwards, “for heaven sake don't neglect it!!!”²⁴ Edwards suggested that eggs be sent in blocks of wood with holes drilled in them²⁵ and larvae be accompanied by “a supply of the hostplant” (Edwards 1873). He kept a large number of vials on hand which he sent to correspondents who offered to supply early stages, asking that they send them back in “a tin box or a cigar box.”^26,27 Edwards abhorred badly preserved inflated (“blown”) larval specimens, like some of those figured by Scudder (1888–1889), which Edwards described as “looking like a coffee sack.”²⁸ Curiously, Edwards readily accepted inflated larvae from Scudder.²⁹ Edwards successfully cultivated a network of many naturalists who provided important specimens for his work.

Edwards reared and documented the early stages of countless species. “When Vol. I. was undertaken,” he explained, “nothing was known by myself or any one else, of eggs, larvae, or chryasalids, except of the more common butterflies…But in 1870 I discovered an infallible way to obtain eggs from the female of any species of butterfly, namely, by confining her with the growing food-plant.” Using this method, he “reared larvae without end” (Edwards 1887–1897). This process was facilitated by the opening of the Chesapeake & Ohio Railroad in 1870, which allowed Edwards to more rapidly convey living early stages to Mary Peart in Philadelphia. Edwards was thereby able to rear numerous species for the first time. He bagged female butterflies and larvae over their hostplants in order to rear them and erected a greenhouse in which to maintain plants during the winter.^30,31 To preserve larvae and pupae during the winter months, he kept some in his bedroom and occasionally sent others to be stored in an ice house located in Clifton Springs, New York.³² He used a similar “ice box” in attempts to induce “ torpidity” in larvae that came from cooler climates.³³ He had trouble, however, rearing species of Hesperiidae, especially “such as pass the winter in larval stage.”³⁴

Edwards grew increasingly obsessed with documenting the biology of the species he figured in BNA. For example, it took nearly thirty years to reveal the entire life history of Parnassius smintheus Doubleday (Edwards 1900–1901). Edwards boasted, “I may venture to say that I have bred butterflies from the egg to a greater extent than any living man no matter who he is or where he lives.”³⁵ Despite his rearing prowess, Edwards displayed little understanding of ecology, a science that was still taking shape during the late 19th century. For example, he “turned loose” near his home about 100 adults of the Old World swallowtail (Papilio machaon L.) which emerged from chrysalids that T. L. Mead had obtained from Germany (Edwards 1882). Edwards caught and released one of them about a week later, but never saw any others.^36,37

Mary Peart also reared many species in order to figure their early stages, thereby contributing a great deal of new biological information. Edwards (1874–1884) wrote, “I have been seconded to the utmost by Mrs. Mary Peart, who has not only drawn the early stages on the stone, but previously on paper, making in each case colored figures; and in order to do this has had to aid in rearing the larvae, and to take a vast amount of trouble upon herself.” Peart meticulously recorded dates of transformations and measurements of the larvae. Edwards once quipped that Peart had contributed more toward his work on butterflies than he had (Walton 1921). After drawing the early stages, Peart preserved representative specimens in alcohol vials which were then sent to Edwards. This added immensely to Edwards' collection of early stages, about which he often boasted. Peart also drew eggs and larval structures with the use of a microscope owned by J. Gibbons Hunt (1826–1893), a prominent Philadelphia physician, botanist, and microscopist who sometimes conducted microscopic work directly for Edwards.³⁸ Probably using Hunt's microscope, Peart rendered illustrations of plant structures for Hunt (1870). On at least one occasion Peart borrowed a microscope from Edwards to draw larval details.³⁹ She later used a solar microscope to examine adult and larval morphology.^40,41,42

Mary Peart prepared colored pattern plates that the colorists used to properly tint the resulting prints.⁴³ The prints of the first and second volumes were colored by Lavinia (Lydia) Bowen (nee Davis) (1820–1888), with the assistance of her sister, Patience Davis Leslie (1826–1893). Although imprints on many plates of BNA indicate that they were colored by L. Bowen, Patience was responsible for a large number. Lavinia Bowen was previously a colorist for John T. Bowen, whom Lavinia married prior to 1838. John Bowen had established the lithographic firm of J. T. Bowen & Company, who published The Viviparous Quadrupeds of North America and the royal octavo edition of The Birds of America, both by John James Audubon. Lavinia and P. D. Leslie reportedly colored prints for these famous works. Lavinia took control of the company after John Bowen's death in 1856 (Barnhill 2010). The female artists and colorists employed by Lavinia's firm worked out of their homes to conceal their service, as domestic affairs were culturally favored for women (Penny 1863). This is supported by Edwards who wrote, “Mrs Bowen the colorist … undertakes to do the coloring for me personally at her house as Miss Peart does the drawing.”⁴⁴ As L. Bowen's health declined, Leslie evidently took over most of the coloring duties for the second volume.

With few exceptions, all the figures on the plates of BNA were hand-colored. Some green larvae, as well as the silver coloring of at least one plate, Argynnis IV (Speyeria coronis Behr), were colored using the process of chromolithography (color lithography).⁴⁵ The printed inscriptions on the plates were done by lettering artists who specialized in this work.⁴⁶ The plates were typically lettered before Edwards wrote the accompanying text, thus he sometimes worked feverishly at the eleventh hour to ensure that he used correct names on the plates. The names on the plates essentially established the nomenclature that he employed in the associated letterpress.

From 1859 until 1867, Lavinia Bowen partnered with J. Cassin (Groce & Wallace 1957), who managed the lithographic work for the firm. During the early production of the first volume, Edwards wrestled with the high production costs of Bowen & Company. “ Cassin charged $30 per plate for the drawing, and 30 cents per sheet for coloring (including paper & printing),” he complained. “The drawing shd be but $15 per plate & coloring not over 20 c[ents].”⁴⁷ Threatening to find other artists, Edwards met with Cassin, who was persuaded to agree to “more reasonable terms.”⁴⁸ After Cassin's death, Bowen took Edward Turnbull as a partner and continued the firm under the name Bowen & Company. However, Edwards was very unhappy with this affiliation, writing, “Since Cassin's death and Turnbull's advent, there has been nothing but misunderstandings. They have altered the prices twice since Cassin's time, and I never know what to expect from them.”⁴⁹ Peart also was dissatisfied, prompting her to inform E. T. Cresson, “I was almost ready to enter a protest against receiving any more work at the hands of Bowen & Co. as that means Turnbull.” She learned that Edwards' work was being badly neglected and suggested that “some other arrangement can be made” (29.xii.1870 ANSP). Apparently, Bowen's personal mail was often intercepted at the firm, prompting her to ask that letters from Edwards be sent to Cresson.

Experiencing ongoing delays in drawing his plates, Edwards scoffed, “I mean to finish the work … if Bowen & Co. do not drag it along for the rest of my life.”⁵⁰ In 1871, Edwards withdrew all pending work from Bowen & Company “on account of a partner in that firm [Turnbull] who is unreliable and of bad habits.”⁵¹ To avoid having to deal with Turnbull in the rendering of his plates, Edwards arranged to work directly with Peart, while the resulting illustrations continued to be printed and colored by Bowen & Company (Edwards 1900–1901; dos Passos 1951). The completed plates were sent to Edwards, who forwarded them to the publisher as needed.

At the conclusion of the first volume, Edwards was relieved: “I feel as if a big thing was off me.”⁵² Its positive reception prompted him to continue publication. Wasting little time, he engaged Peart to begin drawing plates for the second volume, asking that she “keep at it for six months if her health allows & then I will begin to talk of publishing.”⁵³ Although it was his goal to receive 20 plates from Peart before he issued the first part of the second volume,⁵⁴ this plan was not realized. In 1874, Edwards was able to devote more time to the book during a temporary cessation of work at the Kanawha and Ohio Coal Company, which soon after was reorganized (Edwards 1902).

During production of the second volume, Edwards remarked, “It is a fact that I did not foresee the delays in publishing when I began.”⁵⁵ Because of such delays, the publisher suggested that Edwards augment Peart's work with another artist in their employ. Edwards was urged to send specimens for a trial plate, but he worried that this artist would be unable to restore specimens to a natural state prior to illustrating them, as “Miss Peart can do that to perfection.”⁵⁶ Hermann A. Hagen recommended that another artist, Étienne L. Trouvelot (1827–1895), contact Edwards and offer his services for drawing plates.⁵⁷ Trouvelot is inauspiciously recognized for inadvertently introducing the destructive gypsy moth (Lymantria dispar (L.)) into North America in 1869 (Spear 2005). Although Trouvelot completed some illustrations of early stages for BNA, prior commitments prevented him from working on the plates.^58,59 Trouvelot later completed 112 illustrations for Scudder (1888–1889).

The letterpress for plate Chionobas II (Oeneis nevadensis C. Felder & R. Felder), which was issued with Part 3 of the second volume, included wood cut engravings that portrayed the hindwing patterns of four nominal taxa of the genus Chionobas (=Oeneis). This is the only instance where Edwards provided such an illustrated comparison of taxa. Because Edwards' access to relevant publications was limited, he asked P. R. Uhler of Baltimore to copy the original description and provide an outline drawing of the ventral hindwing of Chionobas nevadensis as published by Felder and Felder (1865–1875).⁶⁰ Uhler obliged by “placing the thin paper over the colored lithographs and tracing their details of markings with lead pencil.”⁶¹ Edwards was therefore able to include this taxon in his line-up of Chionobas hindwings.⁶² All the taxa so represented are now generally considered synonymous with Oeneis nevadensis (Pelham 2008). This is an example of the type of assistance that Edwards routinely sought while preparing parts of BNA.

In 1880, Edwards was growing tired of the work. “I am nearing the end of my 2nd vol., and do not propose to continue it,” he told Henry Edwards. “It is a vast trouble and great expense to me.”⁶³ Later the following year, Edwards was still unsure if he would be able continue beyond a second volume.⁶⁴ He confessed that this was dependent upon several things, not the least of which was the cooperation of Mary Peart, without whose help he “certainly would not proceed.”⁶⁵ Edwards remarked, “I have probably 1000 figures by Mrs. Peart in my albums & she is all the time adding to them, so I have the materials for a 3rd vol in hand.”⁶⁶ He boasted that because many of the figured species were rare, no one could rival this collection of illustrations.⁶⁷ He aspired to “live long enough to publish a 3rd volume with all these fine things.”⁶⁸ Edwards proposed that if he could not continue, perhaps “some one hereafter can go on with successive vols by the aid of my drawings.”69

Although the continuation of BNA was in doubt, Edwards remained hopeful: “If I can find the artists I would go on with Vol III & should like to.”⁷⁰ His wishes were realized in 1884, when Houghton Mifflin told Edwards that they were “considering the proposition of a 3rd volume of the Butterflies.”⁷¹ Edwards recommended that he stop the second volume at 50 plates, ending with plate Papilio XIII (early stages of Papilio rutulus Lucas).⁷² Having ended the second volume, he quickly started on the third, vowing to “push it to an end in 4 or 5 years if it is possible”73

While working on new parts for the second and third volumes, Edwards continued to assemble copies of the first volume. Beginning around 1884, he was employing two additional colorists besides Bowen and Leslie to complete outstanding orders for the first volume.⁷⁴ One of these was the art student Mary Caroline “Lina” Beard (1852–1933), a very talented artist in New York who wrote and illustrated several books of her own. It was Houghton, Mifflin & Co. who previously suggested that Edwards contact her, “a lady equal to doing my coloring … daughter of the great artist Beard [James H. Beard].”⁷⁵ Edwards noted that Beard could “only work mornings as she is studying at the League” [Cooper Union and Art Students League in New York City].⁷⁶ With her younger sister, Adelia, Lina co-founded the Girl Pioneers of America, a precursor to the Camp Fire Girls and Girl Scouts of America (Haverstock et al. 2000). Youth organizations led by their brother, Daniel C. Beard, resulted in the formation of the Boy Scouts of America. In addition to coloring plates for new copies of the first volume, Lina Beard also colored half the prints for plate Papilio VII (P. rutulus Lucas) of the second volume.⁷⁷ Beard was a friend of Edwards' youngest daughter, Anne Edwards Smith (nee Anne Scott Edwards) (1858–1930), who sometimes did coloring work for BNA. Edwards mentioned that Anne “ complains of her eyes whenever called to do caterpillars, even my enlarged ones,” yet he proudly proclaimed, “My daughter Anne is my best artist after Mrs Leslie.”⁷⁸ Also employed to work on plates for new copies of the first volume was Shelley W. Denton (1859–1938) of Wellesley, Massachusetts, who also rendered two plates and colored 125 prints of plate Colias II (Colias harfordii H. Edwards) for the third volume.^79,80 Edwards initially described Denton as “one of the best artists I have found”⁸¹ (though he later disliked Denton's lithographic work; see below). In a brief prefatory biography of her father, Vanessa Denton mentioned her father's contribution to BNA (Denton 1949).

Drawing and coloring plates for the third volume proved problematic from the start. During the mid-1880s, Lavinia Bowen's health continued to decline, forcing the very elderly P. D. Leslie to take on more work herself, to “get on as fast as possible.”⁸² Bowen was then doing no coloring for Edwards, but was occasionally able to “aid Mrs. Leslie by her advice as to making colors.”⁸³ In 1887, Edwards reported that Leslie too was “getting old & does much less than she used [to],” adding, “Mrs. Bowen is infirm & nervous to a degree that interferes greatly with Mrs. Leslie's work.”⁸⁴ The sisters lived together and had no servants, thus Leslie did the housework and cared for Bowen in addition to coloring plates for Edwards.⁸⁵ Bowen's death in 1888, coupled with Leslies' declining abilities, made it more difficult for Edwards to complete the plates for the third volume. Edwards lamented, “when I lose Mrs Leslie I shall feel it badly.”⁸⁶ Edwards therefore employed others for this purpose, but he found it “ exceedingly hard to get the coloring done.” He explained, “Most of the ladies who do it take it up but temporarily … they have been art students in New York generally so that I am constantly having to search for a colorist. … These inexperienced persons are sure they can do my work & ruin no end of plates for me before they all [are] convinced that this work is not their vocation.”^87,88 In early 1888, Edwards indicated that he had “4 good colorists whose work I can depend on.”89 Evidently, they were Leslie, Beard, Denton, and Edwards' daughter, Anne.

Edwards devised a plan to use multiple artists to draw plates for the third volume in order to “proceed a great deal faster” than he had with the second.⁹⁰ Because S. W. Denton was also a talented lithographic artist who did work for Houghton, Mifflin & Company,⁹¹ Edwards sent him specimens of Colias chrysomelas (=Colias occidentalis chrysomelas H. Edwards), stating, “If that is done well, I ought to give the artist steady work.”⁹² Denton had three brothers, William D. Denton (1865–1923), Robert W. Denton (1868–1959), and Sherman F. Denton (1855–1937), all of whom also studied natural history and collected butterflies. Their younger sister, Carrie Denton (1869–1959), sometimes accompanied her brothers into the field (Anonymous 1956). In addition to his many famous chromolithographs of fish, Sherman is best known for publishing the two-volume book Moths and Butterflies of the United States (Denton 1897–1900). The four brothers collaborated on the booklet The Butterfly Hunter's Guide (Denton Bros. 1900).

Still in search of additional artists, Edwards sent specimens of two other species, Argynnis lais (=Speyeria hesperis dennisi dos Passos & Grey) and Argynnis nitocris (=Speyeria nokomis nitocris (W. H. Edwards)), to the lithographic firm of J. Sinclair & Company of Philadelphia.⁹³ If he found their work to be acceptable, Edwards desired to retain them for drawing and printing ten plates per year for up to three years.^94,95,96 Edwards, however, was unsure of their ability to do the work as he desired, stating, “they are in the dark & it is proposed that I let them try 2 or 3 plates & then they [can] calculate terms.”⁹⁷ Edwards worried that if they did not accept the work, he would “not know where to find artists.”⁹⁸ He pointed out that Peart could complete six plates per year if not called upon to do larval drawings, which were also necessary for the book.⁹⁹ Edwards hoped to “get a dozen plates ahead before publishing,”¹⁰⁰ but he complained about the cost of doing business with Sinclair & Son. “Sinclair charges so high a price that unless he moderates I shall go to him very little,” he wrote, adding, “His estimate was $75 for drawing alone. I paid Mrs. Peart $25 to $50.”¹⁰¹ Because of Sinclair's high cost, Edwards doubted that he would continue doing business with him.¹⁰² Sinclair's plates would be drawn by Edward A. Ketterer (1860–1909), an artist who lived with his wife, Marie, on South 5th Street in Camden, New Jersey. Edwards described Ketterer as “a fat German fond of his beer.”¹⁰³ Sinclair's firm had previously produced all the plates for Mead (1875), many of which were based on plates and figures that appeared in the first volume of BNA (Argynnis IV, Limenitis II–IV, Satyrus I). Referring to the illustrations in Mead (1875), Edwards noted, “There are 5 or 6 plates of Diurnals, some from my vols,” which were “gotten up under my direction.”^104,105 Unlike the hand-colored originals in BNA, Mead's reproductions were chromolithographed and of substandard quality.

Edwards received the proofs for Denton's first plate, which he initially called, “fairly done.”₁₀₆ Edwards soon after sent specimens of Argynnis liliana (=Speyeria callippe liliana (H. Edwards)) to Denton for a second trial plate.^107,108 Denton evidently employed assistants, as Edwards mentioned that Denton had “lost his labor & proposed at his own risk to do the Argynnis.”¹⁰⁹ Denton “begged to be allowed to try another [plate] on same terms.”¹¹⁰ However, Edwards soon reconsidered the quality of Denton's earlier plate of C. o. chrysomelas and decided not to use it, complaining that the “spots & nerves as well as outlines” were “all more or less wrong.”¹¹¹ Although Edwards was skeptical about Denton's ability to provide a satisfactory illustration (“I fear it will not be accepted”), he initially was pleased with the second attempt, stating, “Denton has succeeded very well with a plate of Arg[ynnis] Liliana of which I got the proof yesterday. Considering the wretched work he made of Colias Chrysomelas it is a wonderful advance.”¹¹² He called the liliana plate “a hundred times better than the other plate.”¹¹³ Although Denton's drawing of liliana had some defects, Edwards considered it a “fair plate & when colored looks well.”¹¹⁴ This drawing was used for plate Argynnis III in Part 3 of the third volume. Denton, however, proved too expensive and Edwards questioned using him for another plate.”¹¹⁵ Denton received no more lithographic business for BNA, though he continued to work as a colorist. He also submitted drawings of larvae to Edwards, but they were likewise unacceptable and “ passed upon by Mrs Peart.”¹¹⁶ In the end, Edwards even replaced Denton's plate of liliana with another by Mary Peart (see Replacement Plates, below).

Edwards was still confident that he would find another artist, asserting, “I will push along even if I go to Mintern Bros. of London for half the plates.”^117,118 Suggested to Edwards by the English entomologist A. G. Butler, the lithographic firm of Mintern Bros. offered to do the work for about $20 per plate, much less than what Sinclair charged, and “in better style” than Edwards could obtain in this country.^119,120 Encouraged by S. F. Baird, Edwards thought that Mintern could complete up to 15 plates.¹²¹ He considered asking Butler or another English entomologist, William F. Kirby, “to superintend each plate, say for £2 per plate perhaps.”¹²² Despite these offers, Edwards never employed Mintern Bros. because of his nagging fear of losing specimens during transit.¹²³

Houghton, Mifflin & Co. again suggested that Edwards send specimens to “a man proposed by them.”¹²⁴ This was Charles Armstrong (1836–1906), an artist and lithographer affiliated with Riverside Press, who operated a firm under the name of Armstrong & Company in Cambridge, Massachusetts (Last 2005). Edwards sent specimens to Armstrong, but upon receiving the trial plate Edwards wrote, “Mrs. Peart utterly condemns the plate … every one of the figures is out of drawing [poorly drawn], wings misshapen, antennae of irregular length, bodies awful … nerves wrong. It is as bad as the first plate Denton drew.”¹²⁵ Edwards considered Armstrong's trial plate to be “ shockingly” poor.¹²⁶ Not surprisingly, Armstrong was not engaged for BNA. An undated letter to S. H. Scudder, probably from the mid-1880s, indicates that Edwards also attempted a trial plate with the firm of O. H. Bailey & Company of Cambridge, Massachusetts. Oakley H. Bailey (1843–1947) is best known for producing hand-colored bird's-eye view lithographic maps of many U.S. cities. Edwards described the results of Bailey's plate as “all out of drawing, not one figure was decently done.”¹²⁷

In March of 1886, Edwards visited Sinclair & Son to see Ketterer's work first-hand. He observed that Ketterer was working on two Argynnis plates, which “ seemed excellent,” noting that “Mrs. Peart approves what she has seen.”¹²⁸ Two months later, Mary Peart informed Edwards that the trial plates by Ketterer for the third volume were good, but she had to “correct them a good deal.”^129,130 Edwards insisted, “In time after half a dozen plates, Ketterer would do satisfactory work, without oversight.”¹³¹ He cautioned, however, that Ketterer knew “nothing of butterflies & the points that are most important.”¹³² Edwards accepted Ketterer's first efforts and they served as plates Argynnis I and Argynnis II in Part 1 of the third volume. Nearly a year later, Edwards noted that Peart “has not yet got Ketterer up to the higher mark, but he improves.”¹³³ Later that year, Edwards stated that Peart's supervision was affording Ketterer “good training.”¹³⁴

Edwards next sent specimens of C. o. chrysomelas to the lithographic firm of J. Bien & Company of New York to replace the failed attempt at this plate by Denton. Bien, however, mistakenly colored their plate using chromolithography rather than hand-coloring.^135,136 Although Edwards initially accepted the results, he reconsidered, declaring that they “made such a bad work of things that I can't use it & did not pay for it.”¹³⁷ Edwards disliked the coloring and compared the legs to those “of a shrimp!”¹³⁸ He argued that it was impossible “ to catch delicate color with chromo.”^139,140 He sent the plate back twice to have it corrected, along with more specimens, but both times it came back without perceptible improvements, thus he “refused the whole.”¹⁴¹Moreover, Bien's artist mistakenly left the box open in which Edwards' specimens were kept, allowing cockroaches to damage some of them. “It was no use to scold,” he wrote, “but I privately lamented.”¹⁴² Edwards finally decided to send specimens of C. o. chrysomelas to Ketterer at Sinclair's firm,¹⁴³ resulting in a plate that was “beautifully done.” Ketterer's illustration was therefore used for plate Colias IV in Part 4 of the third volume.^144,145 Because of her declining health, Peart initially declined to undertake any plates for the third volume,¹⁴⁶ but her devotion to Edwards induced her to stay involved (see below). Edwards liked the work provided by Sinclair, thus most other plates of the third volume were rendered by Ketterer under the supervision of Mary Peart, who was pleased with Ketterer's skills.^147,148 Peart worked on the composition of the plates, reviewed the proofs, and arranged for coloring.^149,150 Edwards was happy that Peart relieved him of “all trouble in that department.” The figures of the early stages in Ketterer's plates were based on colored drawings by Peart, some of which were derived from sketches sent by Edwards' correspondents. Although Edwards worried about Ketterer's ability to accurately reproduce Peart's drawings, he conceded, “I suppose however there is no help for it.”¹⁵¹ Peart allowed Edwards to examine her drawings of early stages before she turned them over to Ketterer.¹⁵²

The first part of the third volume sold very quickly, with one half of its copies exhausted by March of 1887.¹⁵³ Edwards announced, “The publishers wrote last week that Part 1 was going off well and that they thought there was to be a larger sale than for Vol. 2.”¹⁵⁴ He humorously added, “I hope to live to end of vol. 3 & go out, if go I must, in a blaze of Entomol[ogical] glory.”¹⁵⁵ Early the following year, Edwards boasted, “ there never has been such an enquiry for [BNA] as now.”¹⁵⁶ He soon became less optimistic, observing that although sales slowly increased, he was “behind all the time with the artists … and in debt.”¹⁵⁷ This was especially upsetting, as he “put heart & soul” into the book's production.158

Edwards hoped to complete the third volume “inside 5 years,”¹⁵⁹ but it actually took twice that long. The extended period required to produce the book resulted in additional problems. “When I began Vol. 3 it was considerably with different subscribers from what I had with Vol. 2,” Edwards recalled in 1888. “Many of the old ones, in the 20 years since Vol 1, had died. Others found it impracticable to continue the volumes.”¹⁶⁰ By that time, the number of subscribers had fallen from 160 to about 112, yet Edwards was still selling up to 15 sets of the first two volumes annually, mostly to libraries. Edwards wished that more libraries would order his book, noting, “There are enough large Libraries in this country to absorb all the editions of the Butterflies.”¹⁶¹ Fortunately, the book was selling well overseas, particularly in Germany.^162,163

In July 1890, Edwards calculated the number of bound copies of the first and second volumes that had been sold, exclusive of subscriptions. This totaled 174 copies of the first volume and 66 of the second. He had less than 100 plates remaining for the first volume and claimed that it was still being sold ten years after the completion of the second volume.¹⁶⁴ In all, Edwards estimated that by 1890 he had produced 390 copies of the first volume, which included those sent to subscribers. In 1892, Edwards noted an increase in the sale of the first two volumes for reasons unknown.¹⁶⁵

Bound copies of the first and second volumes of BNA were being offered during the 1880s and early 1890s for $35–$45. Seemingly inexpensive, this is equivalent to over $1,000 in our present economy. Individual parts of the third volume were still being issued at a price of $2.25–$3.50 each (HMC 1884). Volumes and separate parts were offered with colored or uncolored (plain) plates, the latter at a substantial discount. All three volumes were available in comparable formats for many years (RP 1899). Edwards often wrote about sending plates to the publisher for the assembly of additional sets. Over a period of about 60 days, he could arrange to color enough plates to assemble seven copies of the second volume.¹⁶⁶ Nonetheless, coloring plates for new copies of the first two volumes interfered with work on the third volume.¹⁶⁷ In 1888, new bound sets of the first volume were offered for $30 each.168

As the third volume progressed, Ketterer became increasingly difficult and unreliable, leading Edwards to lament, “Ketterer is [so] infernally slow with the plates that I am afraid I may be an Octn [octogenarian] before I get thro [vol.] 3.”¹⁶⁹ Edwards learned from Mary Peart that Ketterer was an orchestra leader in Camden, New Jersey, which Edwards suspected accounted for “the great delays in my plates.”¹⁷⁰ Edwards needed to rely on Ketterer, as Peart was engaged in “making pencil drawings” and “supervising the execution of the plates on the stone.”¹⁷¹ In early 1893, Ketterer wrote that he was giving up every other piece of work and intended to devote himself to plates for BNA, completing two every “ do no such thing,” adding, “It looks as if he was in earnest, but he is so giving to promising & not performing that I can't tell.”^172,173 True to form, Ketterer soon informed Edwards that he was delayed because of “ somebody's death.”¹⁷⁴ Despite this aggravation (“Ketterer vexes me out of years growth”), Edwards admitted that Ketterer's plates were “good” when they finally arrived.¹⁷⁵ As consolation, Sinclair proposed that Edwards did not have to pay for any plate until it was drawn and printed.176 Despite this agreement, Ketterer did not improve, prompting Edwards to condemn Sinclair's “intolerable slowness.”177

Sinclair evidently suffered from financial difficulties.¹⁷⁸ In 1889, he sold his business to George S. Harris & Sons, who through this purchase became the largest lithographic firm in Philadelphia (Piola 2012). Edwards was thankful that “all the artists go along … and they promise great care in my works.”¹⁷⁹ Optimistic of this new relationship, Edwards wrote, “This firm promises to do my work as well as Sinclair's & I hope they will.”¹⁸⁰ Soon after, however, Edwards was again complaining of delays and errors, noting that two initial proofs for Part 7 of the third volume were “badly printed,” calling them “worse than any I ever had from Sinclair.”¹⁸¹ Edwards remained hopeful that “they may however do better.”182

Well into the third volume, coloring plates remained Edwards' biggest headache. He “tried and rejected” many colorists during the production of the third volume.¹⁸³ He admitted that he suffered “great worry and trouble over colorists,” explaining, “they work a few months & get used to it & change for something that pays better. Then I have to search again and try several perhaps before I find one who can & will do such work carefully.”¹⁸⁴ Edwards mentioned the possibility of employing a “Miss Grace Stend,”¹⁸⁵ but I have been unable to confirm that she was retained.

In 1890, Edwards began work with a colorist named Mary Ann “Donalda” Downie (1860–1894) (WE journal “ W” WVSA), who lived in Winnipeg, Manitoba, Canada. Edwards described her as “one of the best artists I have had.”¹⁸⁶ The last 55 prints that she completed were for plate Chionobas IX (Oeneis spp.) in Part 15 of the third volume. Edwards received the finished plates on 11 April 1894. Downie was spending the winter of 1893/94 in Mexico City, where she enjoyed “time and plenty of light.” Among Edwards' manuscripts (WVSA) is an express receipt for a shipment of “Estampas” (pictures, i.e. printed plates) to Mexico, dated 19 March 1894. This suggests that Edwards sent her another set of plates, but they remained unfinished due to Downie's untimely death on 7 May 1894.

In 1893, Edwards considered the idea of advertising for colorists in New York.¹⁸⁷ Not long after, he learned that P. D. Leslie was too feeble to continue working on his plates, thus ending their long relationship.¹⁸⁸ A few weeks later, he heard that Leslie was “affected in the head and can attend to nothing”¹⁸⁹ (she died soon after). Edwards continued to fill the void with other colorists, including Clara H. Dutton (1857-?) of Wolcott, New York, who worked on at least one plate, that of Satyrus III (Cercyonis oetus charon (W. H. Edwards)). After coloring 75 prints of this plate for Part 16, she confessed, “I have never felt so tired from an order as from these … since they are finished I really feel rested.”¹⁹⁰ Another colorist who apparently worked on plates for BNA was Mary C. Drew (1866-?), daughter of Thomas B. Drew, who was the first curator and librarian of the Pilgrim Museum in Plymouth, Massachusetts. Edwards described her as an “expert colorist.”¹⁹¹ Edwards' wife, Katherine T. Edwards (1827–1901), also colored plates for BNA (Edwards 1902), but her involvement was likely minimal. Edwards often shared coloring jobs among different colorists, allowing the work to progress much more quickly. This was done for the very last plate of the third volume.¹⁹²

The colorists often provided detailed information to Edwards about how they painted the plates. Edwards copied some of this information into his personal journals so it could be passed on to other colorists. For the figures of S. diana in the first volume, P. D. Leslie advised, “In Diana the black is made of indigo blue, carmine, & enough gamboge to change the color.” Carmine was a bright red pigment, while gamboge was mustard yellow. Referring to the figures of Argynnis lais (=Speyeria hesperis dennisi dos Passos & Grey) in the third volume, Leslie related, “We have always used gamboge & carmine. These flies look as tho' they would require a little dull color. We should use a little umber carefully put on, as there is danger of getting these colors too dark. If when done, they are too light, a little pale umber should be passed over them, that is better than having them too dark at first … it is best to take an old plate to try the colors on and work them up as strong as wanted.” The colorists usually “sized” the plates with a solution to prevent the pigment from penetrating too deeply into the paper.¹⁹³ Because some colorists “spoiled as many [plates] in sizing as they did in coloring,” Edwards sized the prints himself before sending them to those who were inexperienced in this method.¹⁹⁴

Certain colors were especially difficult to reproduce. For example, silver powder (“silver saucer”), used to create silver markings on figures, often turned black and required correction.¹⁹⁵ Edwards remarked, “I have found nothing that will wear as silver on my plates. I have consulted artists & Mrs. Bowen knows of nothing.” He resolved to “leave the white of the paper,” noting that the effect was “a great deal better than tarnished silver.”¹⁹⁶ He solved this problem in 1886 when he discovered that aluminum powder could be mixed with water and applied with a brush.¹⁹⁷ Edwards also observed that the green colors being used for larvae were insufficient.¹⁹⁸ At the urging of Peart and Ketterer, Edwards decided to use chromolithographic coloring for greens, which was then “touched with brush afterwards.”¹⁹⁹ This method was apparently used for several plates of satyrid butterflies of the third volume (e.g. plates Satyrodes I and Neonympha II). The cost of doing this was no more than a standard plate.200

The artist Lillie Sullivan (1855–1903), who long served as the Chief Illustrator in entomology for the U.S. Department of Agriculture, contributed some drawings for the third volume. She rendered some illustrations of larvae (WE journal “S” WVSA) and her figures of adult Oeneis macounii (W. H. Edwards), which were originally completed for the Canadian entomologist James Fletcher, were reproduced by Ketterer for plate Chionobas X.²⁰¹

In early 1893, Edwards noted that he had on hand twelve bound copies of the first two volumes and 100 parts of the third. In addition, he had about 150 uncolored copies of each plate from the first volume and 175 of each from the second. Edwards wanted to keep the colorists busy during down time in the production of the third volume. He remarked, “I am compelled to give the ladies work [in] order to keep them.”²⁰² By early 1894, however, new sales were tapering off. Edwards informed S. H. Scudder that the publisher had sold fewer parts of the book, and only two bound copies, during the previous six months.²⁰³ Three years later, Edwards said that Houghton Mifflin possessed six sets of the first volume, two of the second, and 15 of the third. Although the preparation of additional copies of the first two volumes was very timeconsuming, ²⁰⁴ they continued to be assembled long after their original parts were issued. These copies combined newly colored plates with remaining letterpress. In 1895, Edwards anticipated a call for further copies of the first two volumes for subscribers who “held back till the [third] volume was completed.”²⁰⁵ Edwards planned to sell copies of the third volume for $50 each.²⁰⁶

Approaching the end. Edwards often remarked that the creation of BNA absorbed all of his spare time.^1,2,3 “If I had known what I had to go thro when I began the Butt. N. A.,” he wrote, “I never should have made the beginnings. It grew on me as I proceeded.” The second volume was not anticipated, but once started it dominated all his “summers and winters” for ten years. When he reached the third volume, he suspected that he would be “tied by the leg at least 8 years from the difficulty in getting work done.”⁴ Edwards thought that he could keep going with the third volume, maybe issuing as many as 100 plates, “if everything favored.”5

Nearly 30 years after proposing to publish illustrations of North American butterflies, Edwards thought he would “never get to the end.”⁶ He remained optimistic about the possibility of a fourth volume, which Scudder urged him to consider.⁷ William J. Holland previously told Edwards about a “new process” which would facilitate the production of color plates, supposedly to encourage Edwards to continue on to a fourth volume.⁸ This was possibly a reference to color photographic reproduction. In the end, it was Holland himself who used this less expensive method for the plates in his widely popular The Butterfly Book (Holland 1898). By making his book much more affordable, Holland ushered in a new era of Lepidoptera study in North America. Instead of fearing that Holland's book would render his own obsolete, Edwards confidently predicted that its success would boost sales of his own book. “It seems to me the larger the sales of this book the better for both of us,” he advised Scudder. “Many will become interested who will push these studies farther & more for our books.”⁹ Holland disclosed to Edwards that he had to sell 8000 copies of The Butterfly Book to see any financial return.¹⁰ Greatly exceeding this goal, the book was reprinted until 1931, when it was revised and reprinted until at least 1949.

Forging ahead, Edwards visited Ketterer in May 1895 and discovered that he was also doing work for the lithographic firm Ketterlinus Printing House of Philadelphia, who was engaged in a large job to draw bird eggs for the second volume of Life Histories of North American Birds (Bendire 1895).¹¹ Later that year, Edwards asked Ketterer to complete one more plate, which was not yet done by March of 1896. Edwards complained that Ketterer had been working on it for four or five months, “tho at any time he could have drawn the plate in one week.”¹²

Discouraged, Edwards wrote in 1896 that Ketterer was still “fooling with” his last plate.¹³ Edwards trusted Peart, but his confidence in Ketterer was still shaken. Ketterer repeatedly told Edwards that the last plate was nearly drawn, yet he was not working on it: “He works for the great lith[ographic] houses & slights my work.”¹⁴ At his wits end, Edwards maintained that if it had not been for Ketterer's “lies and delays,” the third volume would have been done three years earlier.¹⁵ By this time, Edwards regretted ever doing business with Ketterer and avowed that if he continued with a fourth volume, he would not utilize the artist.^16,17 To complicate matters, Edwards discovered that W. J. Holland retained Ketterer to draw an illustration for Holland (1896).¹⁸ This frustrated Edwards, who thought Ketterer should be concentrating on completing his last plate for BNA. Ketterer assured Edwards that the final plate for the third volume would be completed “in 2 or 3 days.”¹⁹ Edwards finally received the proofs of this plate four weeks later, a full seven months after Ketterer supposedly began work on it.²⁰ Between May 1895 and May 1896, Ketterer completed only this single plate for BNA.²¹ As a result of these delays, Mary Peart agreed to complete plates Parnassius I and Chionobas XIII (Edwards 1887–1897).

Edwards desired to devote at least one entire plate in the third volume to a swallowtail butterfly which he and the artist-naturalist David Bruce had collected in 1894 in Glenwood Springs, Colorado.²² Edwards had previously described this species as Papilio brucei (=Papilio machaon bairdii Edwards). He hoped that he could raise the extra money to produce an additional part of the book to accommodate this plan. To this end, Scudder tried to obtain a grant of $125 from the Elizabeth Thompson Fund.²³ Presumably this money did not come through, as the plate was never published. Nonetheless, Edwards issued eight pages of text on P. brucei in the last part (18) of BNA.

The third volume was completed in early 1897, when Edwards sent the remaining text to the printer. “[T]hat ends the matter,” he wrote, “Glad to get this printing business off my hands.”²⁴ Although he was thankful to be done with the third volume, Edwards admitted, “I shall take a rest, but if I see hereafter that I can begin another series of Plates, I mean to have them drawn.”²⁵ He envisioned issuing no more than “about a dozen” additional plates.²⁶ The following year, Edwards informed Scudder, “I have so many species of rare larvae to figure that I can't stop … if I have the money to go farther.”²⁷ Although Edwards surmised that he could “go to 60 plates comfortably,” he was concerned that so many would make the third volume too bulky. He proposed to “cut it into 2 vols, 30 plates each!”²⁸ Excited by this possibility, he wrote, “It seems to me that is a good idea.”²⁹

Unfortunately, events precluded the enlargement of the third volume or the production of additional volumes. Edwards later recalled, “I was urged to go on with the fourth volume, for which I had on hand a great deal of material, both insects and drawings. … But as I was well advanced in years, I felt that I had better stop now” (Edwards 1900–1901). Although Edwards once entertained the idea of passing the production of BNA on to his son-in-law, Theodore L. Mead (“I hope you may take up the Butterflies of NA where I leave it, or continue it”30), the project came to an abrupt end.

Sales of the book had slowed considerably by 1899. In six months, only two copies of the third volume were sold, and none of the others.³¹ Edwards wrote about “a great lot of superfluous plates of the 3 vols.” stored in the attic and various other rooms of his home. Disappointed, and perhaps somewhat resentful that he could not continue his book, he considered the remaining plates to be a nuisance and threatened to “ burn all but 50 sets.”³² Edwards believed they were of value only to him: “After I go to Kingdom Come there is no one who can do anything with my plates.”³³ He presumably gave many of these extra plates to Mead, who in 1922 asked the Philadelphia entomologist Henry Skinner for advice on how to dispose of them.³⁴ Skinner suggested that he contact the entomological bookseller John D. Sherman, Jr. of Mount Vernon, New York (“I don't know what he would pay as the policy of dealers is to buy cheap and sell dear”). The fate of these plates is unknown, as they were not listed among Sherman's catalogues during the 1920s. However, a set of 93 plates (mostly uncolored) were offered by the same bookseller earlier that same year on behalf of Edwards' daughter, Anne (Sherman 1922). It was noted in Sherman's catalog that all the remaining text of the book had previously been destroyed.

Soon after the completion of the third volume, Edwards' interest in Lepidoptera waned. He instead dedicated his energy to publishing books about the spelling of Shakespear's name and the history of the Edwards family (Edwards 1900–1901; Bethune 1909). The entomologist Charles J. S. Bethune reassured Edwards that his work “will long endure & keep your name in deserved honor during many generations to come.”³⁵ A few years later, Edwards professed, “I stand the years pretty well. Have good appetite & good digestion and sleep the sleep of a naturalist.” He amused himself by working in his flower garden and trimming his shrubbery. His last publication on butterflies appeared in 1898, only a year after finishing BNA: “I have let them fly since the end of my vol 3.”³⁶

Financial obstacles. Not long after embarking on BNA, Edwards became extremely concerned about its cost, writing in 1868, “I am troubled about the expense … to get my drawings made and colored.”¹ The following year, Edwards disclosed that he lost up to 20 cents in commission on each part of the book that was sold through booksellers.² Bowen & Co. went so far as to suggest to E. T. Cresson that he consider coloring the plates using chromolithography to “much lessen the cost of this fine work.”³ In 1871, Edwards revealed that he had lost $3,000 on the publication of the first volume.⁴ According to an expense sheet maintained by E. T. Cresson (ANSP), Edwards personally paid at least $1,714.50 to see the first volume published. This amounts to about $31,000 today. The first volume grossed the American Entomological Society only $20 (about $369 today). Expenses included type for the letterpress, binding, lithography, coloring, and ink, as well as postage for mailing proofs, plates, and lithographic stones. Edwards was very self-conscious about his expenses for the first volume, confessing that he “burned all the bills so that no one should see how extravagant I had been.”^5,6 In 1879, Edwards observed that it was “very expensive publishing these things … I am out of pocket with every Part I issue.” He explained that he could take no less than one dollar per plate to cover his costs in drawing and coloring the plates, as well as printing the text.”I should hate to do the work in a cheaper manner with second rate artists,” he wrote.⁷ He revealed that it took every cent he could raise to enable him to finish the book.^8,9 He was forced to increase the price per full volume from $30 to $40, but this only covered expenses.¹⁰ Resolute, he declared, “I have any way to publish the Butterflies, money or none!”¹¹

Although Edwards called work on BNA “a labor of love,”¹² it became increasingly more difficult to publish the parts and he ran short of funds after the completion of the second volume. “I desire exceedingly to proceed with Vol. 3,” he wrote in 1885, “but am restrained for lack of money to meet the payments to artists &c.” Edwards admitted that he had “sunk a great sum of money in publishing the 2 vols,” but was “willing to pay out another considerable sum to get Vol 3 done.”¹³ Edwards hoped that the production of the third volume would “run itself” if he reached sufficient sales of the parts, but this proved disappointing. Before any parts of the third volume were published, Edwards confessed, “ Vol III is costing me more to get started than I had calculated.”¹⁴

To conserve funds during the production of the third volume, Ketterer produced two drawings on one lithographic stone, which limited the choice of species that Edwards could use for each plate. He was forced to “ select two [species] of somewhat equal degree of darkness and light,” thus he often had to re-arrange his planned order of species.¹⁵ At the beginning of the third volume, Edwards recalled that his first outlay was $800 before he even received any drawings for the first two parts.¹⁶ Fortunately, Houghton, Mifflin & Co. had great faith in the book and did not force Edwards to pay arrears. He referred to the firm as “remarkably kind people.”¹⁷ In early 1895, Edwards was indebted to the publishers for $600.¹⁸ The following year, Edwards was afraid to ask Houghton Mifflin to continue covering his expenses, as he was “on their books at least $700 now.” Still self-conscious of his situation, he asked Scudder to “ say no more about it.”¹⁹

According to Edwards' writings, lithographic work cost $20–50 per plate. Because colorists were required to complete multiple copies of each plate, their pay could be five times greater than that of the lithographic artists. Coloring expenses for the first and second volumes were typically 25–50 cents per sheet ($5–$12 today), yet he paid up to 75 cents for more complex plates, such as Lycaena II of the second volume.^20,21,22 Edwards lost money on this plate, as he charged subscribers only 50 cents each.²³ Edwards worried what Lavinia Bowen would charge to color the complex plate Lycaena III of the second volume: “I have not yet learned and I don't want to for all that matter!”²⁴ Coloring costs for the third volume remained the same.²⁵ Up to 450 copies of each plate were struck before the lithographic stones were cleaned to produce other plates.^26,27

Something that Edwards always regretted was using the stereotype method to print his text. “I made a great mistake in having the text stereotyped”, he recalled, “ this cost me $3000 for the 3 vols.”²⁸ Printing the text cost about 3.5 cents per sheet. He believed that it would have been cheaper to strike a large number (i.e. 500) of typeset parts at one time, rather than stereotyping small amounts over an extended period. Edwards considered stereotyping to be a “useless expense.”²⁹

Costs were not just associated with production. During the publication of each volume, Edwards provided gratis copies to a number of people, including E. T. Cresson, H. Edwards, and S. H. Scudder. Referring to Cresson, Edwards later admitted, “I do send Cresson the Parts as they appear & have always done so for the reason that he helped me so much with vol. 1 & got little for his pains.”³⁰ In addition to her pay, Edwards gave Mary Peart free copies of each part of the book (WE journal “D” WVSA). He also provided a copy of the first volume to Samuel F. B. Morse, who coinvented the single-wire telegraph system and the original Morse code signal language (Morse called the plates “exceedingly beautiful”³¹). Edwards also granted “ special exceptions” to any lepidopterists who applied for a payment plan.³² Though generous and politically savvy, these practices were fiscally unwise.

Edwards' collection. To help reduce his financial burden and allow him to produce a third volume of BNA, Edwards settled on a notion to make his collection “subservient to the book.”¹ In 1885 Edwards struck a deal to sell his collection to W. J. Holland for $2,500 (equivalent to about $58,000 today). Holland, who in 1884 had purchased the collection of Edwards' son-in-law, T. L. Mead, desired to unite these important collections. Although Edwards described his collection as “the most complete N.A. collection of diurnals that ever has been made or probably ever will be,” it comprised only 4,417 adult specimens. Considered fairly large at that time, this is a small collection by today's standards (some present-day private collections contain well over one million specimens). Edwards also had a great number of early stages preserved in alcohol vials, noting that there were “many rare species represented & altogether such a collection is not likely to be made again.”²

Edwards “started down the butterfly path” in 1856 when he lived in New Hamburgh, New York. His collection “increased largely” shortly after moving to Newburgh, New York in 1859 (Edwards 1900–1901). “I may expect to bring together a large Collection of Lepidoptera in no long time,” he eagerly announced.³ Edwards credited the Lepidoptera catalog by Morris (1862) with helping him “know whether a given insect had been named and described or not” (Edwards 1900–1901).

In 1870, Edwards expressed an interest to bequeath his collection to the Museum of Comparative Zoology (Harvard University),^4,5 but his relationship with H. A. Hagen subsequently deteriorated. He claimed that Hagen was not “admirably decent” and declared “as the Museum is now managed I will not send them a fly.”⁶ After his falling-out with Hagen, Edwards thought his collection would probably go to the Smithsonian Institution, where he hoped it would be kept intact and not “stolen piece meal.”⁷ He did not trust public collections, fearing that thieves would eventually remove the best specimens.⁸ He also distrusted certain other lepidopterists, such as Herman Strecker, who was long suspected of illicitly removing specimens from various collections that he visited⁹ (McClain et al. 2002; Leach 2013). Although Holland (1909, 1928) maintained that Edwards contemplated selling his collection to the British Museum, I found no firm offers to that affect among Edwards' correspondence. Edwards' letters clearly indicate that his collection was formerly promised to MCZ.¹⁰ Holland possibly wished to conceal the fact that Edwards' collection was intended for another domestic repository to avoid making his own acquisition appear self-serving. “I have always pleased myself with the thought that I was rendering a service to the cause of American science by retaining in this country Mr. Edwards's types,” he wrote, “ I think I ought to have credit for doing what I did” (Holland 1909). Of course, this was written after Edwards' death.

Edwards did not publicize the sale of his collection for some time, even keeping it from his wife and children.¹¹ Edwards asked Holland to consider the sale confidential, but he doubted that Holland could keep it a secret.¹² Although the sale helped to pay the bills of the third volume, Edwards was depressed about it, recalling the following year, “It was painful to part with the collection. … It was [like] pulling eye teeth.”¹³

Holland paid Edwards $500 up front and committed $1000 towards publishing expenses for each half of the third volume. This was Edwards' idea to ensure that he would not spend the money too quickly. Edwards received Holland's first payment on 14 January 1886, and Edwards shipped the first specimens of Hesperiidae in cigar boxes two weeks later.¹⁴ Specimens of Lycaenidae followed, along with other specimens that Edwards did not need for the third volume. Sending extra specimens allowed Edwards to “ have room to display better” what remained.¹⁵ Their agreement did not include any specimens that Edwards would collect after the sale.¹⁶

Edwards continued to ship portions of his collection to Holland for many years and regularly advised Holland when payments were due. He often asked him to send bank drafts directly to those requiring payment (colorists, printers, etc.). On several occasions, Edwards admonished Holland for being late with payments and once reprimanded him by quoting a section of their contract: “Holland agrees to forward to Edwards drafts as for the amount from time to time required, on receiving notice thereof from said Edwards.”¹⁷ In the end, Edwards did not use the initial $500 for the third volume, but rather for “clearing all arrears on Vols 2 & 1.”^{18 Still dejected about the extreme impact of BNA on his finances, he lamented, “It has nearly squelched me … this and business matters here make life not worth living. It certainly would not be had I not Entomological mental relief.”19} By late 1888, Holland had paid “all he agreed to & something over.”²⁰

Although Holland's payments enabled the continuation of BNA, Edwards also obtained funds from other sources. He received an unsolicited grant of $500 from the National Academy of Science (Edwards 1887–1897), which Scudder had arranged without his knowledge.²¹ Additional grants totaling $350 were received from the Elizabeth Thompson Science Fund (Edwards 1887–1897; Anonymous 1889). Edwards also applied profits from the continuing sale of the first two volumes toward the publication of the third, noting, “it takes all these to pay for vol 3.”^22,23 He later confirmed that all the proceeds from the sales of the first two volumes (“every cent”) went into producing the third volume.24

Shipments of specimens to Holland continued at least through April 1897, when Edwards wrote, “I am packing up the greater part of the butterflies remaining to go to Dr. Holland.” Although he would miss having the specimens, Edwards confessed that it was “a relief to get rid of the collection.”²⁵ By that time, he had surely grown tired of its maintenance and it probably served as a constant reminder of his inability to realize his dream of producing a comprehensive book on all the North American butterflies.

Publication particulars. The title pages issued at the completion of each volume were dated as follows: 1868–1872 (vol. 1), 1884 (vol. 2) and 1897 (vol. 3). It is very difficult, however, to ascertain the true dates of issue for the associated parts of each volume. Holland (1928) attempted to clarify some dates of issue, but he relied almost entirely on the dates as they appeared on the original wrappers for each part. Evidence indicates that some wrapper dates preceded the actual dates of issue by as much as five months.

I discovered that Edwards rarely recorded the actual dates of issue. For the second volume, he received this information from Scudder, who had asked a librarian at the Boston Society of Natural History (BSNH) to record the dates as they received each part from the publisher, who was located in nearby Cambridge, Massachusetts. Edwards then applied these dates for his “Dates of Issue” page, which was included with the last part of the volume. In 1896, Edwards again asked Scudder to obtain the dates of receipt for the third volume. He admitted, “I thought I regularly entered them as they issued, but I find I did so to 7 [parts] out of the 17.”1 Edwards was more confident of his data for parts of the third volume, deciding to list specific dates instead of just the month and year as he had previously done. Edwards provided the dates of issue to the publishers, who evidently did not record this information themselves.

I reviewed published and unpublished evidence related to the dates of issue for parts of BNA, which was then compared against the dates given on the title pages of the individual parts, as well as those listed on the “ Dates of Issue” sheets for each volume. This information is presented in Tables 2–4. The literature contains many additional references to the dates of issue, but most are invalid estimates only. Published library donation records are helpful, but they do not embrace all the parts. I include only those references which appear to best indicate the actual dates of issue.

Edwards often struggled with forthcoming issue dates in his correspondence, complaining on numerous occasions about the delays of various parts. A publisher's “Notice”, printed on the wrapper of Part 5 of the first volume, explains the reason for many of these problems. This part was issued in April of 1870, but the wrapper reflected the date when the text was printed, namely December 1869 (Table 2).

Such interruptions were not uncommon during the production of all three volumes. A trip to Europe by Mary Peart in 1871 resulted in a delay in completing Part 9 of the first volume.2 Edwards worried that another delay in issuing this part was due to a small pox epidemic in Philadelphia.³ Corrections to the lithographic stones and coloring sometimes stalled production.^4,5 A fire at the publisher's in Cambridge in 1877 caused a loss of 50 completed parts that were ready for distribution⁶ (a second fire at the same location in 1879 did not damage any materials for BNA7). In 1886, a labor strike in Philadelphia delayed the printing of the first two plates for the third volume.⁸ As the third volume was coming to a close, the last two plates were held up after the elderly printer fell down in the street and bruised himself so badly that he was unable to work for several days.⁹

In his Preface for the second volume, Edwards (1874–1884) attributed further delays to the inclusion of “much original matter on the early stages” and the associated “labor of preparing and coloring the Plates.” Edwards told Holland that these delays were “caused by the inability of Mrs. Peart to draw the figures on stone any faster,” yet he was “unwilling to employ any other person on that part of the work.”¹⁰ In 1886, Edwards remarked that Peart “is much of an invalid at present & I can't count at all on active aid.”¹¹ Two years later, Edwards again noted that delays were due to Peart's infirmity, revealing, “She has always been an invalid and ended up work on my plates of Vol. 2 as the body would permit.”¹² The difficulty in retaining skilled colorists after the loss of Bowen and Leslie also slowed publication.¹³ In addition, production was affected by natural events, such as flooding during the spring of 1886.¹⁴ Probably in an attempt to conceal such interruptions, most parts of the second volume and all parts of the third bore only the year on the wrappers. Some parts were issued in portions, thereby allowing completed copies to be distributed, followed by the remainder of the copies at a later date.^15,16 Paid subscribers were given priority over those receiving free copies.

Table 2.

Issuance of parts for the first series (volume 1) of BNA.

A supplemental part for the first volume was issued in January of 1873. Seventeen pages in length, it included the title page for the volume, indexes, and eleven replacement pages for the Synopsis. Also included was a new “Dates of Issue” page on which an entry for the supplement was inserted. As the supplement was being assembled, the new publisher (Hurd & Houghton) decided to offer 40 leather-bound sets of the first volume for $30 each, “for the Holidays.”^17,18,19,20 They distributed a one-page advertisement announcing that a “ limited edition has been printed … complete in itself and independent of any subsequent ones.” However, like many of the individual parts of the book, it appears that this edition was delayed. Although Edwards asked Lavinia Bowen to color 50 additional sets, he estimated that this work would take at least two or three months and feared this was “too slow for anything of a demand.”²¹ A year later, Edwards remarked that Bowen “ shall soon have 50 more sets ready,” but this seemingly relates to Edwards' earlier request, which was still not completed.²² During the spring of 1874, Hurd & Houghton asked E. T. Cresson, who had originally arranged the publication of the first volume, to send them the electrotype plates for the text in order to “ print a new edition” (24.iv.1874 ANSP). The 50 additional sets of plates were presumably used to assemble copies of the “limited edition”, which probably were not available until mid-1874; well over a year after the edition was proposed. Unlike previous copies of the first volume, this edition included a revised title page designed like that of the original wrappers, bearing the longer title The Butterflies of North America with Colored Drawings and Descriptions. It was imprinted with the name of the new publisher, Hurd and Houghton, with a date of 1874. Its Synopsis also possessed a new title page, dated 1874. Although Holland (1828) suggested that only new title pages were printed in 1874, evidence indicates that the entire text was reprinted and combined with newly colored plates.

Prior to the commencement of the second volume, Edwards issued a lengthy “Notice”, dated November 1872, in which he mentions the planned availability of the first series of parts as a bound volume (i.e. the “ limited edition”). Printed on the rear wrapper of the supplemental part was the following:

Although Edwards indicated that the first part of the second volume would issue “about 1 June 1873,” it actually appeared a year later (Table 3). This is an excellent example of how difficult it was for Edwards and the publisher to accurately forecast dates of publication.

In 1879, Edwards observed that no copies of the first volume had been available “in the past year.”²³ He remarked that the publisher “wrote to ask how many copies of the text they should print of my vol 1 for the new Plates that I am having colored.”²⁴ He instructed them to print another 100 copies of text, but asked that they add the year 1879 to the title page, which was reconfigured by Edwards. After using a longer title for the 1874 reissue, Edwards reverted back to Butterflies of North America on the advice of H. A. Hagen.²⁵ This reissue, published by Houghton, Osgood & Co., also included the revised “Dates of Issue” sheet and a corrected “Alphabetical Index”. As noted by Hemming (1931), a copy of the first volume which Edwards presented in 1879 to the Entomological Society of London (now Royal Entomological Society of London) possesses penciled corrections to these pages. I examined images of these corrections and found that they were possibly written by Edwards himself. Annotations on the “Dates of Issue” sheet are consistent with the dates given in the 1879 reprinting, of which Hemming (1931) was apparently unaware. In the original “Alphabetical Index” there was an error in citing the letterpress for Parnassius eversmanni [Ménétriés] which began on page 27, not page 25 as initially indicated.

By 1881, reprinted text for the first volume was again exhausted. Edwards mentioned that “a number of new parties” wanted the first two volumes, but they were unavailable. He informed Holland, who requested an extra copy of the first volume, that he would receive it “ within a year.”²⁶ In 1882, Edwards realized that there was a greater demand than he could possibly keep up with.²⁷ New copies of the first volume were assembled whenever possible, but only when the colorists were “ out of work” with the second volume.”²⁸ Due to the continued demand, text for the first volume was reprinted once again in 1888, this time by Houghton Mifflin. It was titled like the 1879 reissue and incorporated the same corrections.

During the production of the second volume, Edwards complained about errors in several parts. He discovered two mistakes in the lettering of the plates and noted that the printers “had made 3 blunders in printing Parts of vol 2 so that the genera in these cases could not come together.”²⁹ He corrected these errors where possible, thus enabling subscribers to have “ Plates & pages bound correctly.”³⁰ In addition, the 16 pages of letterpress that accompanied plates Lycaena II and Lycaena III of Part 12 were mistakenly paginated by the printer. This was never corrected and it remains the only portion of the book with printed page numbers.

As the second volume was nearing completion, Edwards issued a “Notice To Subscribers” with Part 12, dated 1 May 1884 (despite the reference to a second plate in Part 13, it was not issued):

Part 13, issued in January of 1885, also included the title page, preface, supplementary notes, and a list of species (see below).

Table 3.

Issuance of parts for the second series (volume 2) of BNA.

In late 1885, Edwards instructed Mary Peart to begin working on plates for the third volume, but she had a backlog of other drawings to finish.³¹ As a result, Edwards did not receive completed copies of the first two plates until April of 1886.³² Subscribers later received a lengthy “Advertisement,” for the third volume, dated 25 December 1886. A copy of this document was date-stamped by the Smithsonian Institution as having been received one day before the first part of the third volume, suggesting that it was mailed separately. In the advertisement, Edwards stated that he envisioned publishing ten plates per year, and “ probably three or four will contain figures of eggs exclusively.” He also suggested that it may be desirable to include sixty plates in the third volume instead of fifty. None of these plans were realized. Rather than ten plates appearing each year, an average of only 5 plates was issued each year during the volume's ten-year production between 1887 and 1897. After 1891, only three plates were issued in most years, and none were issued in 1896. Instead of 60 plates, only 51 were completed, and none exclusively portrayed eggs. Because he had so many drawings of eggs, Edwards had previously proposed issuing them in a separate book, which never transpired.³³

During production of the third volume, Edwards discovered an error on the proof for plate Coenonympha I, on which the lettering artist misspelled the plate name as “Ceonympha”. Edwards attempted to correct the error, but the printer had already struck all 450 prints. Edwards took the advice of Scudder and resolved to “take the plates as they were, noting the error in the text.”^34,35 A similar error previously affected plate Anthocharis I of the second volume, on which the species name Olympia was misspelled as “Olimpia”.

Part 10 of the third volume was delayed due to “bad work on the third plate, which required that it be reprinted.”³⁶ In addition, two pages of Part 15 had to be reprinted and reissued in Part 17. Various corrections to lithographic stones and to the lettering on the plates delayed the issuance of other parts.³⁷ Part 6 was previously delayed due to the artist (Ketterer) using one lithographic stone to complete two plates, one of which was for Part 7.³⁸ Ketterer also made various mistakes on his drawings which Peart worked to correct.³⁹ Ketterer sometimes struggled with accurately portraying details, such as eggs or the legs of adult butterflies.⁴⁰

A separate publisher's notice, dated February 1897, was issued with Part 17. It referred to the printer's errors and indicated that an 18th part would be issued. The publication of this additional part is not widely known, as it was not included on the “Dates of Issue” sheet for that volume. The notice read:

Edwards sent the last proof plates for the third volume to the publisher in November 1896, after which he started preparing the title page, preface, supplements and index.⁴¹ Part 17 was issued during early March of 1897. Part 18, which also included text (but no plates) for Papilio IV and Papilio V, was issued during late May of 1897.

Two versions of the title page for the third volume were issued. One featured the ornament (emblem) of Houghton Mifflin, while the other bore the ornament of the printer, Riverside Press. The latter ornament was created in 1885 (RP 1899), after the completion of the second volume, on which the Houghton Mifflin ornament appeared. Because the Houghton Mifflin ornament was used for the title page of the second volume, as well as that of the 1888 reissue of the first volume, a similar title page was probably preferred for later copies of the third volume. This consistency was important when all three volumes were sold as a set.

Replacement plates. In 1871, when Edwards was nearly done with the first volume, he decided to redraw three plates. He had recently received new specimens of Argynnis leto (=Speyeria cybele leto (Behr)), which convinced him that he needed to improve plate Arygnnis X of 1869 (Fig. 3). He initially thought the best way to do this was to simply ask the colorist, P. D. Leslie, to recolor the female on the original plate to conform to the new specimens.^1,2 Unfortunately, the cost of doing this was prohibitive (ten cents more than a standard plate), thus Edwards resolved to have Mary Peart redraw the plate³ (Fig. 4). The following year, Edwards received male and female specimens of Speyeria nokomis which were collected in 1871 in Arizona. Because Daniel Wiest's original plate of the species (Fig. 5), issued in 1868 as Argynnis IV, was drawn from a single male specimen in poor condition, Edwards instructed Peart to create a new illustration (Fig. 6).^4,5 Upon sending an Arizona specimen of the species to S. F. Baird in 1872, Edwards wrote, “The Argynnis in the first box is really Nokomis, figd from a poor specimen (and unique till now) in Part 1 … probably I will figure both sexes & make a new plate for [Part] 1.”⁶ Although Skinner (1918) claimed that the original plate of S. nokomis was not published, it obviously was sent to subscribers.

Figs. 3–6.

Original plates (left) and replacement plates for BNA. 3, Argynnis X by M. Peart. 4, Argynnis X by M. Peart. 5, Argynnis IV by D. Wiest. 6, Argynnis IV by M. Peart.

Unfortunately, Edwards' replacement of the plate of S. nokomis resulted in some later confusion, as the taxon portrayed on the original plate differs from that depicted in the replacement (Skinner 1918; Holland 1928; Brown 1965, 1983). Variation in the coloration of various copies of the original plate heightened this confusion (Holland 1928). The source of the lost holotype originally figured by Edwards is uncertain (Brown 1965). Its condition was poor, thus Edwards likely did not recognize differences between it and those from Arizona.

Edwards was also forced to create a new plate for Speyeria diana, which had originally been drawn by D. Weist for plate Argynnis I, issued in 1868 (Fig. 7). Although Edwards had previously described this plate as “handsome,” he now realized that “the plate as it stood was such as to make coloring doubly expensive.”7 As with the other replacements, the new diana plate was drawn by Mary Peart (Fig. 8).

Edwards intended to issue all three new plates with Part 10 of the first volume, but this would have delayed its publication. He therefore decided to include them in the supplemental part, issued in January 1873. In a “ Notice To Subscribers” printed on the wrapper for Part 10, Edwards explained,

The new plates of S. nokomis and S. leto were accompanied by updated letterpress. Upon the publication of the new plates, Edwards wrote, “Those three butterflies are the finest in North America and Miss Peart and Mrs. Bowen are equal to portraying them. It is a satisfaction to work with such artists.”⁸ Referring to the supplement, Edwards asserted, “if you don't say it contains three remarkable species I give up.”9

Edwards obviously possessed an extreme fondness for S. diana, which he considered to be the most beautiful species of its genus, if not all butterflies. He saw his first live male of this species in 1864 and wrote an account of finding a female a few days later, which was thought to be the first female known to science (Edwards 1868–1872) (per Strecker (1900), a female had actually been found in Missouri around the year 1853). It even seems that Edwards commissioned a song, the “Diana March”, in its honor.¹⁰ Edwards claimed that S. diana became “scarce” in the vicinity of his home after 1865, when two visiting collectors took about 100 specimens (dos Passos 1951). In 1873, he told Herman Strecker, “Either Mr. Ridings [James Ridings] or the fires in the woods have extinguished the breed hereabouts.”¹¹ It seems, however, that the butterfly had not become as rare as Edwards maintained, as he continued to record observations of adults in the area for decades thereafter.

It has long been known that the first volume contained three replacement plates (e.g. Holland 1928; dos Passos 1951), but another replacement went unnoticed. In 1887, shortly after S. W. Denton's plate of Speyeria c. liliana was issued as Argynnis III in Part 3 of the third volume, Edwards began to express his dissatisfaction with its quality, noting that it was “not very well drawn nor very well printed.”¹² He resolved, “ Probably before end of the volume I will have the Liliana plate re-drawn & sent out gratis.”¹³ Edwards had argued with the firm who had struck the 450 prints from Denton's liliana plate. They were printed on poor paper and the impressions were not as clear as the proof which Edwards had approved. In addition, Edwards was displeased with the coloring, noting that “It ended with Denton scraping somewhat the dark parts.”¹⁴ Edwards identified the printing firm as “Morse & Co.,” possibly referring to George H. Morse, Jr., a Boston printer whom Denton presumably hired to strike his plates for Edwards. Because of these problems, Morse reprinted the entire lot of this plate at his expense,¹⁵ yet Edwards still regarded these prints as unacceptable.

Edwards ultimately replaced Denton's liliana plate (Fig. 9) with another drawn by Mary Peart (Fig. 10). There are very few references to the new liliana plate among Edwards' letters. In August 1897, he told W. G. Wright, “As to Liliana, Mrs. Peart began a plate of this on stone from the fine example you sent me last year. When it will be finished I do not know, but sometime this fall probably.”¹⁶ When he later compared the new plate with the original version, he was “more disgusted with the latter than ever.”¹⁷ In the letterpress for the original plate of liliana, Edwards (1887–1897) remarked, “By an oversight, the egg and young larva were not figured on the present Plate, but will be given on Plate V of this series of Argynnis.” In fact, the early stages of liliana did not appear on Argynnis V, which was issued in Part 9, but instead were later incorporated into Peart's new version of the liliana plate (Fig. 10). After the book was completed, Edwards remarked that he was finally able to pay Peart for drawing the Liliana plate, “for which the beneficent Gods be thanked.”¹⁸ Probably referring to the new liliana plate, Edwards expressed his regret in late 1898 that Scudder “did not get one of the new prints.”¹⁹ Because of its belated appearance, the new liliana plate was apparently not sent to all the original subscribers and instead was incorporated into later bound copies of the volume. This explains why Edwards (1887–1897) indicated that only two plates of the third volume were drawn by Peart (Parnassius I and Chionobas XIII), while some existing copies of the volume (including my own) have an additional plate that bears her name. This replacement plate is perhaps even more rare than the three replaced in the first volume.

Figs. 7–10.

Original plates (left) and replacement plates for BNA. 7, Argynnis I by D. Wiest. 8, Argynnis I by M. Peart. 9, Argynnis III by S. W. Denton. 10, Argynnis III by M. Peart.

Other plates received modification, but were not completely redone. Edwards was sometimes unhappy with the initial results and ordered new pattern plates be created to correct early attempts. For example, the male figure on plate Chionobas II (Oeneis nevadensis gigas Butler) of the second volume was derived “not from life,” but from a drawing of the holotype in the British Museum.²⁰ The drawing was sent to Edwards by A. G. Butler, who had previously described the taxon as Oeneis gigas. Edwards considered the original coloration of this figure on plate Chionobas II to be inaccurate. In 1891, the San Francisco lepidopterist W. G. Wright collected additional specimens of the species in British Columbia, which Edwards used to correct new copies of the plate.²¹ In addition to such corrections to prepared plates, the pattern plates were sometimes “renewed” to prevent the colorists from migrating away from the original colors.²² Replication of previous errors was often a problem during the production of hand-colored illustrations.

Synopsis. In 1867, S. F. Baird suggested that Edwards increase demand of his book by including a “ list of all described N. Am. species of the different genera figured and even Synopsis.”¹ Edwards agreed, but jokingly asked, “Should I adopt all of the new genera of Tom, Dick & Harry, or have I discretion to do as I think best about that?”²

The publication of the Synopsis of North American Butterflies, which accompanied the first volume of BNA, was very complex. Its production reveals that Edwards was struggling to keep up with a rapidly changing taxonomy. When considering the usage of names, he sometimes asked the advice of other entomologists, such as H. Edwards³ and T. L. Mead. In his personal journal of 1871 (see Calhoun 2010), Mead mentioned helping to proofread parts of the Synopsis while he was visiting Edwards in Coalburg.

The Synopsis was issued with Parts 3–10 of the first volume of BNA. Pages five and six, issued in Part 4, were replaced with Part 5 less than eight months later. With Part 10 came additional changes, when Edwards revised at least seven more pages (1–4, 10, 11, 19). I obtained copies of the original nine pages that Edwards issued in five separate parts between December 1868 and January 1871. These pages came from a copy of BNA once owned by the lepidopterist Paddy B. McHenry, who published a brief historical summary of the Synopsis (McHenry 1952). These pages reveal that Edwards changed the format of some citations and altered the arrangement of many taxa to reflect more current research. Upon ordering an extra copy of the entire Synopsis, T. L. Mead was advised by E. T. Cresson that “there have been many corrections,” thus Mead's order was held until all the corrected sheets could be sent within the supplemental part of the first volume (27.ix.1872 WVSA). Perhaps somewhat sardonically, Mead later called the Synopsis “very interesting” (TM journal entry, 9.x.1871). The original version of the Synopsis, as issued in installments, is very rare.

Probably in December 1872, the updated Synopsis was separately offered for sale by the American Entomological Society in the same size (quarto) as BNA. After receiving the separately-published version, the botanist Lewis R. Gibbs considered it to be the “ authoritative one,” in preference over the version sent in installments.⁴ In early 1873, after the first volume was concluded and the Synopsis had been separately issued, Edwards discovered yet another error, the omission of the species Gonepteryx clorinde (=Anteos clorinde (Godart)).⁵ This page was corrected and the new version was included in newly assembled copies of the first volume. Although the Synopsis was considered helpful, the generic arrangement and meager citations were criticized ([Scudder] 1874). The entomologist John Hamilton considered the Synopsis to be “valuable to those who have acquired some proficiency, but of little account to the beginner.”⁶

Revised Synopsis. About the year 1877, Edwards decided to issue a revised Synopsis for the second volume of BNA. At that time, Edwards was preparing the more comprehensive Catalogue of Lepidoptera of America, north of Mexico (Edwards 1877). He believed that such a catalogue was necessary because “a large number of new species have been described, belonging to the North American fauna.” Edwards intended to base the revised Synopsis in BNA on his 1877 catalogue. He asked Henry Edwards to review proofs of his catalogue, whose comments would also apply to the Synopsis.¹ The revised Synopsis was meant to update the original version by “adding only reference to species named or found to be N. Amn since 1872.”² Edwards characterized the new Synopsis in 1882 as “barely a list except that all species names since 1872 … will have references.”³ He added, “I don't intend this to take the place of a catalogue for general use and I propose in course of a year to issue a new edition of my Cat[alogue] of 1877 for sale.”⁴ (This was not actually published until 1885.)

Virtually unknown, this publication was entitled Synopsis of North American Butterflies. Revised and Brought Down to 1882. It was begun in Part 10 of the second volume of BNA, yet it was plagued with problems from the start and never completed. After receiving a letter from H. A. Hagen, who had identified several errors, Edwards bemoaned the fact that this Synopsis lacked two species through “omissions in copying,” and the printer had forgotten to paginate the first installment.^5,6 Edwards was also uncertain about the status of several taxa and felt that he needed to correct those treatments. Justifiably frustrated, Edwards wrote, “I should not have issued this Synopsis now had I not promised to give it … it will cost me $100, perhaps $150, and I shall get no return.”⁷

The first (and only) installment of this Synopsis was only eight pages in length. It contained species numbers 1–143, beginning with Papilio Ajax (=Eurytides marcellus (Cramer)) and ending with Argynnis Polaris (=Boloria polaris (Boisduval)). Edwards continued to work on it, noting that the publisher was setting type “ from Erebia to and into Lycaena.”⁸ Like the Synopsis that accompanied the first volume of BNA, Edwards intended to include a preface, lists of authors and abbreviations, addenda, and corrigenda.⁹ He made a few changes to the sheets he had already printed, adding page numbers and replacing some species names.^10,11 However, Edwards soon questioned the continuation of the revised Synopsis and decided to suppress the pages he had already printed, planning to “rewrite as complete a Catalogue as I possibly can & issue it as Part XIII” of the second volume.¹² He soon reversed this decision as well, doubting the logic of even including such a compendium in the book. After asking for advice from H. A. Hagen, Edwards decided to discontinue the revised Synopsis and prepare a much more detailed second edition of his 1877 catalogue for publication elsewhere.¹³ He explained, “I had my intention to carry the Syn[opsis] to 1883 adding new matter but I will make it as perfect as I can now” [as a separate catalogue].¹⁴ He also noted that the inclusion of a detailed catalogue in BNA would have increased the size of the second volume, making it “too bulky.” Still wanting to include some kind of summary in the second volume, Edwards reconciled this dilemma by deciding to issue in Part XIII “a List of species with no references at all—a bare list.”¹⁵ The revised Synopsis purportedly reached 26 pages,¹⁶ but no more were distributed beyond the eight in Part 10.

Subscribers were notified of the discontinuation of this Synopsis in a notice dated 1 April 1882, which was issued with Part 11 of the second volume:

Edwards recalled copies of the revised Synopsis, thus very few have survived. I examined images of a copy in the Ernst Mayr Library (Harvard University), which bears an inscription at the top of the first page: “This Synopsis of vol. II was stopped 1882 on my suggestion, to be replaced by a new one.” The handwriting matches that of H. A. Hagen, who donated his personal library to MCZ. While working at Harvard College (later Harvard University), Hagen retained his books in his office, along with all others on entomology belonging to the College (Winsor 1880 ). Hagen subscribed to BNA and was a regular correspondent of Edwards for many years. This evidence confirms that it was indeed Hagen who dissuaded Edwards from completing the revised Synopsis.

List of Species. After Edwards abandoned the revised Synopsis for the second volume, he decided to include a basic list only, “for the reason that many subscribers to Vol 1 probably know of no change in arrangement since Vol 1 issued” and such a list “would show that an entire change had been made, or many superfluous genera rejected.”¹ Entitled List of Species of the Diurnal Lepidoptera of America North of Mexico, it included its own title page, dated 1884. Introductory comments by Edwards were dated 1 November 1884. This list was issued with the last (13th) part of BNA in January of 1885. This list was merely intended as a compendium of the names and varieties presented in Edwards' updated catalogue (Edwards [1885]). The List of Diurnal Lepidoptera was separately offered for sale by Houghton Mifflin for $0.50 each and could be ordered from the publisher or Edwards himself.

Complete citations. Based on this study, I suggest the following citations for The Butterflies of North America and related publications. This is the first time that such comprehensive citations have been offered for the book. Citations for the third volume include only the month of issue, as the specific days given on the “Date of Issue” sheet for that volume (Table 4) are the dates of receipt by BSNH, not the actual dates of publication. These citations essentially follow the bibliographic format of Pelham (2008).

EDWARDS, WILLIAM HENRY

1868–[1873]. The butterflies of North America [with colored drawings and descriptions]. Philadelphia: American Entomological Society 1: (1) ii, [63–76], pls. [20–24] (Apr [Jun] 1868), (2) [41–44, 77–80, 135–136], pls. [12, 13, 25, 26, 45] (Aug [Oct] 1868), (3) [81–84, 127, 128, 141–144, 149–152], pls. [27, 28, 41, 47, 49] (Dec 1868 [May 1869]), (4) [45–52, 85, 86, 129, 130, 145–148], pls. [14, 15, 29, 42, 48] (Apr [Sept] 1869), (5) [53, 54, 87, 88, 99, 100, 131, 132, 153–156], pls. [16, 30, 35, 43, 50] (Dec 1869 [Apr 1870]), (6) [37, 38, 55, 56, 89–92, 133, 134], pls. [10, 17, 31, 32, 44] (Jun [Aug] 1870), (7) [17–20, 39, 40, 57, 58, 97, 98, 137–140] pls. [4, 11, 18, 34, 46] (Jan [Mar] 1871]), (8) [29–36, 93–96, 101–110], pls. [8, 9, 33, 36, 37] (Aug 1871), (9) [1–16, 111–120], pls. [1, 2, 3, 38, 39] (Dec 1871 [Jan 1872]), (10) [21–28, 59–62, 121–126], pls. [5, 6, 7, 19, 40] (Jul [Sept] 1872), (Suppl.) t.p, [ii], [4 pp.], [73, 74, 85, 86, 157–164], pls. [20, 23, 29; all replacements] (1872 [Jan 1873]).

[1869]–[1873]. Synopsis of North American butterflies. In: W. H. Edwards, The butterflies of North America[with colored drawings and descriptions]. Philadelphia: American Entomological Society 1: (3) [1]–4 (Dec 1868 [May 1869]), (4) 5–6 (Apr [Sept] 1869), (5) 5–6 (Dec 1869 [Apr 1870]), (6) 7–14 (Jun [Aug] 1870), (7) 15–22 (Jan [Mar] 1871), (8) 23–38 (Aug 1871), (10) 1–6, 9–12, 19, 20, 39–50 (Jul [Sep] 1872), (Suppl.) t.p., v, 51–52 (1872 [Jan 1873]).

1872. Synopsis of North American butterflies. Philadelphia: American Entomological Society. vi+52 pp.

1874. The butterflies of North America with colored drawings and descriptions. New York: Hurd & Houghton; first reissue of vol. 1: [ii]+[4]+[164]+vi+52 pp, 50 pls. {1874}.

1874–1885. The butterflies of North America [with colored drawings and descriptions]. New York: Hurd & Houghton 2: (1) [1–4, 77–82, 137, 138, 275–278, 289–294], pls. [1, 16, 26, 43, 46] (May 1874), (2) [5, 6, 87–92, 141, 142, 183–188, 305–310], pls. [2, 18, 28, 33, 48] ([Oct] 1874), (3) [25–28, 131, 132, 231–244, 279–284, 311–314], pls. [6, 24, 38, 44, 49] ([Jun] 1875), (4) [33–38, 129, 130, 133–136, 151–160, 189–192], pls. [8, 23, 25, 30, 34] (Nov [Dec] 1875]), (5) [93–102, 139, 140, 143–150, 245–258], pls. [19, 20, 27, 29, 39] (Jul [Sept] 1876]), (6) [7–24, 51, 52, 259, 260], pls. [3–5, 12, 40] ([Dec] 1877); Boston: Houghton, Osgood & Co. (7) [43–46, 83–86, 103–116, 161–182], pls. [10, 17, 21, 31, 32] ([Dec] 1878), (8) [193–200, 201–220, 285–288], pls. [36, 36, 45] ([Dec] 1879]); Boston: Houghton, Mifflin & Co. (9) [29–32, 39–42, 47–50], pls. [7, 9, 11] ([Oct] 1880), (10) [117–128, 261–274], pls. [22, 41, 42] ([Jun] 1882), (11) [67–76, 221–230, 295–304], pls. [15, 37, 47] ([Apr] 1883), (12) [53–66, 315–332], pls. [13, 50, 51] ([Jun] 1884), (13) t. p., [6 pp.], [333–338], pl. [14] (1885 [Jan 1885]).

1879. The butterflies of North America. Boston: Houghton, Osgood & Co.; second reissue of vol. 1: [ii]+[4]+[164]+vi+52 pp, 50 pls. {1879}.

1882. Synopsis of North American butterflies: revised and brought down to 1882. In: W. H. Edwards, The butterflies of North America [with colored drawings and descriptions]. Boston: Houghton, Mifflin & Co. 2: (10) 1–8. {Jun 1882} (incomplete).

1884. A List of species of the diurnal Lepidoptera of America north of Mexico. In: W. H. Edwards, The butterflies of North America [with colored drawings and descriptions]. Boston; Houghton, Mifflin & Co. 2: (12) [2 pp.], [12] pp. {Dec 1884}.

1884. A List of species of the diurnal Lepidoptera of America north of Mexico. Boston; Houghton, Mifflin & Co. [2]+[12] pp. {1884}.

1887–1897. The butterflies of North America [with colored drawings and descriptions]. Boston & New York; Houghton, Mifflin & Co. 3: (1) [71–76, 91–94], pls. [7, 10, 11] ([Jan] 1887), (2) [77–82, 97–100, 205–212], pls. [8, 13, 27] ([Apr] 1887), (3) [95, 96, 129–134, 145–148], pls. [12, 18, 21] ([Sept] 1887), (4) [87–90, 135, 136, 219–224], pls. [9, 19, 29] (1887 [Jan 1888]), (5) [149–152, 185–192, 247–252], pls. [22, 25, 34] ([May] 1888), (6) [1, 2, 57–64, 213–218], pls. [1, 5, 28] (1888 [Jan 1889]), (7) [65–70, 245, 246, 253–256], pls. [6, 33, 35] ([Mar] 1889), (8) [3–6, 137–144, 153–174], pls. [2, 20, 23] ([Jun] 1889), (9) [101–108, 225–230, 257–266], pls. [14, 30, 36] ([Feb] 1890), (10) [109–114, 125–128, 193–204], pls. [15, 17, 26] ([Oct] 1890), (11) [125–184, 231–236, 277–290], pls. [24, 31, 38] ([Apr] 1891), (12) [7–14, 293–306], pls. [3, 40, 41] (1891 [Jan 1892]), (13) [291–292, 307–332], pls. [39, 42, 43] ([Dec] 1892), (14) [267–276, 333–340, 361–368], pls. [37, 44, 47] ([Nov] 1893), (15) [115–124, 341–360], pls. [16, 45, 46] ([Jul] 1894), (16) [35–52, 237–244, 369–380], pls. [4, 32, 48] ([Oct] 1895), (17) [53–56, 83–86, 123, 124 (sheet corrected), 341, 342 (sheet corrected), 381–410], pls. [49, 50, 51] ([Mar] 1897), (18) t. p., [8 pp.], [15–34, 411–430] ([May 1897]). Pl. [12; replacement] (?Apr 1898).

Table 4.

Issuance of parts for the third series (volume 3) of BNA.

continued

1888. The butterflies of North America. Boston; Houghton, Mifflin & Co.; third reissue of vol. 1: [ii]+[4]+[164]+vi+52 pp, 50 pls. {1888}.

Portrait of Mary Peart. In 1898, one year after the completion of BNA, Mary Peart's portrait was painted by her niece, the artist Caroline Peart (1870–1963). It represents the only known likeness of Peart, which has never before been published in connection with entomology (Fig. 11). A stylized depiction of the Asian swallowtail butterfly, Teinopalpus imperialis Hope, is hung above her right shoulder in reference to her many years of work with butterflies. Also represented is her deep affection for cats, a passion she shared with Edwards when they visited a cat show together in Philadelphia in 1884: “No end of pretty kittens of all colors.”¹ Rendered in oil on canvas, the portrait measures 105.4 × 80 cm (41.5 × 31.5 in) and is preserved in The Phillips Museum of Art, Franklin & Marshall College, Lancaster, Pennsylvania

Mary Peart was one of several talented artists in her family. Besides her niece, she was the great aunt of the celebrated 20th century Pennsylvania artist, Andrew Wyeth (1917–2009). Wyeth viewed the portrait of Mary in 1963, when it was bequeathed with other paintings by Caroline Peart to Franklin & Marshall College (Lestz 1963).

With the exception of Edwards' brief comments in connection with BNA, very little has been published about Mary Peart. Born in Pennsylvania on 16 April 1837, her parents were Quakers from Chester County (Futhey & Cope 1881). During her younger years she shared a residence in Philadelphia with fellow Quaker Graceanna (Grace Anna) Lewis (1821–1812), a prominent naturalist who is most recognized for her work in ornithology (Bonta 1991). Peart's religious affiliation probably led to her later association with the naturalist and lithographer John Cassin, another Quaker from Chester County, Pennsylvania. Cassin also was acquainted with Lewis and named the white-edged oriole, Icterus graceannae (Icteridae) in her honor. Census records reveal that Peart and Lewis were neighbors in Philadelphia during the 1880s.

Peart was credited as having “no living compeer in her special department of Butterflies” (Anonymous 1902). Her long-time friend, Graceanna Lewis honored her as a fellow woman-scientist, “Gifted with true genius for her art,” someone who had “chosen to devote herself to the illustration of scientific subjects, and has succeeded so well as to prove the fitness of her choice” (Lewis 1874). As a testament to Peart's “extraordinary ability in figuring insects on stone,” six plates from BNA were exhibited in 1876 at the Women's (“Ladies'”) Pavilion at the Centennial Exposition in Philadelphia (Dimmock 1876). Peart was greatly admired by entomologists, including Henry Edwards.² Henry Skinner of Philadelphia characterized her as “a delightful woman; cultured, refined and modest to a high degree” (Walton 1921). Skinner also pronounced her “a wonderful artist,” whose work “was as good as any in the world” (25.ix.1922, Mead corresp., MGCL).

It is obvious that Peart's contributions elevated the quality of BNA far beyond that initially envisioned by Edwards. In honor of her thirty years of dedicated service, Edwards described the butterfly Chionobas peartiae, which is now recognized as a junior subjective synonym of Oeneis polyxenes subhyalina (J. Curtis). Edward Ketterer illustrated this butterfly on plate Chionobas XIV in Part 17 of the third volume of BNA. “ I intend to name the species from my lovely assistant Mrs. Peart,” Edwards revealed to Scudder in 1895.3 Originally proposing the name peartii, he finally settled on peartiae, explaining, “It strikes me that [if] named for a lady the ae would imply as much.”⁴

Edwards credited Peart with rendering 2,500 figures for BNA, “the beauty and precision of which it has not been possible to copy on the lithographic stones” (Edwards 1887–1897) (Fig. 12). Edwards considered Peart to be irreplaceable, both professionally and personally. He stated that she had “that sort of eyes & sees many things that I never should see.”⁵ Edwards obviously held a special place in his heart for Peart and visited her whenever he was in Philadelphia. They continued to correspond for at least a decade after the book was finished.

Figs. 11–12.

Mary Peart and her lithographic work for BNA. 11, portrait of Peart, 1898 (courtesy Franklin & Marshall College, Phillips Museum of Art). 12, studies of eggs and larvae from plate Lycaena III, second volume (at bottom is an enlarged attribution from plate).

During the production of the second volume of BNA in 1876, Edwards was afraid that he would lose Peart's services as an artist. In a panic, he incorporated extensive passages about Peart in several letters, which reflected his high regard for her. “I heard distressing news (to me) that my Miss Peart was to be married some time this season and could do but little more work for me,” he worriedly wrote. “It took all the breath out of me … she has taken a personal interest in the work and troubled herself more than any other one can and will.” Edwards continued his praises at length: “If I find one who can make a passable drawing of a butterfly, I can't hope to get one who will make a tolerable caterpillar. It took Miss P. some time to catch the limits of these butterflies, as you will see of comparing the plates of vol. 1. It was only by taking living butterflies and watching their ways of standing, and of using legs and antennae that she reached the present perfection on this point. She is lovely in character and as gentle in her manners as her work is beautiful. Miss Peart it seems is to marry an Englishman [John S. Peart] of the same name as herself & no relation. It is so remarkable a name that it is one of the oddest things.”^6,7,8 Edwards considered the marriage of the two Pearts to be “a singular coincidence.”⁹

Peart thought that she would relocate abroad after her marriage, thus ending her work for Edwards. Although Edwards immediately sought contingencies (“I have heard of a young lady in Phila and I hope to induce her to take hold”¹⁰), the work on BNA would have been irrevocably impacted following the loss of Peart's artistic talent, not to mention her expert rearing abilities. To Edwards' relief, Peart soon assured him that she would stay in Philadelphia after her marriage and continue to do work for him, at least occasionally.¹¹ Peart married on 28 May 1877 and moved from 533 North 4th Street to 422 Wetherill Street (now S. Carlisle St.) in Philadelphia. Edwards visited them at their “nice little house” a few weeks after their marriage.¹² He described her husband as “an intelligent man English by birth but has no accent and seems to have lived a good deal at Cape [of] Good Hope.”¹³ Edwards, however, still worried that “the petty cares of housekeeping on a limited scale will fritter away the time of Mrs. Peart.”¹⁴ As it turned out, Peart continued to work with Edwards for another two decades.

Peart's husband died unexpectedly in January of 1889. Troubled by this event, Edwards stated that Peart was “ too afflicted to make drawings. … She is so delicate that there is always cause for apprehension lest she too may suddenly fail me in which case the Butterflies [BNA] would be in a bad way.”^15,16 Peart worked on BNA during the most productive time of her life, between the ages of 31 and 60. In addition to the many plates for BNA, she rendered illustrations for various articles, including Edwards (1878, 1883).¹⁷

Regrettably, Edwards did not have the opportunity to publish most of Peart's drawings of eggs, larvae, and chrysalids. He loaned them in 1887 to S. H. Scudder,¹⁸ who published chromolithographs of some for his book (Scudder 1888–1889). Many of the same figures on Scudder's plates were later reproduced by Holland (1898, 1931) and Klots (1951). Among Scudder's correspondence at BMS is Edwards' list of these drawings, dated November 1887, which includes Scudder's annotations. Edwards worried about the wellbeing of these drawings and hoped that no harm would come to them while in Scudder's hands.¹⁹ He remarked that Peart's drawings among Scudder's plates can always be recognized by their “accuracy & beauty of outline or position.”²⁰ Edwards requested that Scudder return the originals in short order. “I really need the drawings for pigment reference in coloring,” he wrote, “they are very valuable.”²¹ Having not received them five months later, Edwards chided Scudder, “If you have done with my drawings send them home soon.”²²

After the completion of BNA, Edwards wondered what to do with all of Peart's unused renderings. “I wish I could dispose of my drawings,” he wrote, “much by Mrs Peart—fully 2500 figures, arranged in 8 or 10 albums.” Peart needed the money and had an “equal interest” in their sale.”²³ Edwards wished to see them deposited in a public institution where they would be available for reference (Lyman 1900). He asked Scudder if he knew of a buyer, stating, “They are too beautiful to be lost sight of.”²⁴ Some of these drawings are now preserved at WVSA.

Peart's frail physical condition was an ongoing concern to Edwards. In 1875, he noted that Peart's health “would not permit her working at drawing at present.”²⁵ A decade later, he reported, “Mrs Peart is so great an invalid that I am never sure when she will be able to make drawings,” revealing that she was “confined to her bed a great deal.”^26,27 Peart downplayed a particular bout of illness as an “attack of indisposition.”²⁸ In 1889, a Mrs. L. S. Johnson informed Edwards that Peart was unable to write: “I am writing for her. … She will write more particularly when she can use her hand”²⁹ Edwards described this affliction as a “swollen fore finger on [her] right hand.”³⁰ By February 1890, Peart's hand was healing and she could again hold a pen.³¹ After the completion of BNA, she joyfully announced, “I get out of the house on good days and call myself well.”³²

During the last years of her marriage, Peart lived at 1901 Vine Street in Philadelphia, “just across the square” from The Academy of Natural Sciences.³³ After the death of her husband, she resided for a time with her sister, Rebecca Bean, in Schuylkill, Chester County, Pennsylvania. She also lived briefly in Pawling, Chester County, Pennsylvania,³⁴ where she had spent her summers during the 1870s (20.viii.1873, Mead corresp., RC).

Probably as a result of the many years she spent rearing larvae for Edwards, Mary developed an interest in botany and during the 1890s became a member of the Botanical Society of Pennsylvania (Harshberger 1899). Around 1898 she moved to 113 N. Woodstock Street, where she boarded with the Joseph F. Bamback family. Mary died on 12 April 1917 in the Homeopathic Hospital of Pottstown, Pennsylvania at the age of 80. She was buried next to her husband in Morris Cemetery, Phoenixville, Chester County, Pennsylvania. Peart's invaluable contributions to BNA are a testament to her unrelenting dedication to her very good friend and benefactor, William Henry Edwards.