Help for Evaluating Prints

This page provides extra information on the following topics:

Want more than just the basics?  The learning area (when completed) will have more information (including photos if available) on a specific topic.


An obvious sign that a print is a reproduction is the presence of writing that should not be there.  Suspect would be credits on the front or the back that acknowledge private companies, print manufacturers, libraries, universities, museums, or non-profit organizations such as the Audubon Society.  If it says, "From the Collection of...," "Reproduced from...," "Copyright by...," "Courtesy of...," or "Limited Edition," it may be a very nice print, but it is definitely not an Audubon original.

All original Audubon plates have legends and other writing including subscription part number, plate number, artist and printer credits, plus a title.  Other credits may be seen in the form of initials or full credit lines, particularly on octavo plates.  Although legends and other writing may be trimmed away, it usually is not. Since almost all reproductions include the original writing from the print, we think it is a mistake to spend a lot of time transcribing writing, rather than putting your efforts elsewhere.  If you are looking at a bird print, then note the printer credit on the bottom right. This will tell you whether your print (or reproduction) is from the Havell Edition, the Bien Edition, or is instead an octavo print (Bowen or Endicott).  If you are looking at an animal print, Bowen is the only relevant printer for the folios, while Bowen or Nagel & Weingaertner are credited with the octavos.  

Return to TOP

Sheet Size

Each Audubon edition has a certain size of paper associated with it.  Because binding was not uniform, there is a great deal of variation in the exact sizes of prints that are considered untrimmed.  Do not worry about small variations. The original double elephant folio sheets of the Havell Edition measured 39-1/2 inches by 26-1/2 inches (a typical "full sheet" print might be about 38 x 25).  The paper used in the Bien edition was a similar size, 39-1/2 inches x 26-1/2 inches.  Imperial Folio quadrupeds were printed on paper that measured 28 inches by 22 inches (a plate that is "full sheet" would be around 27 x 21 or better).  Bird octavos, when disbound, are usually in the neighborhood of 6-1/2 inches by 10 inches, while quadruped octavos are often a little wider at 7 inches x 10 inches.  Size alone cannot establish the authenticity of a print, but sheet or image size can often be used to quickly disprove authenticity.

Return to TOP

Image Size 

The images in the Havell and Bien editions come in various sizes depending on the birds being depicted and the way Audubon chose to portray them. All birds were painted life size. As a consequence, some plates are populated with small birds that take up only a small percentage of the available area, whereas other plates show large birds that are depicted in contorted positions so they will fit on the page.  Because the backgrounds used in the Havell and Bien editions vary for some prints, image sizes are not always consistent between these editions, but bird sizes should be.  If you don't have a good grasp of bird sizes, do what Audubon expert Bill Steiner suggests -- look the birds up in a field guide.

Most Imperial Folio quadruped plates show the animals in a landscape.  We once spent a wonderful afternoon looking through the Library of Congress's three volumes of the folio, and our impression was that most images, even those of very small animals, tended to fill a good portion of the available space.  One happy side effect of this approach is that trimming is not as much of a problem for this series as it is for the Havell Edition.

Return to TOP

Color Painted (Not Printed)

Except for Biens, the major editions were colored by hand.  Most reproduction prints (including plates from the Audubon facsimile Amsterdam and Abbeville Editions) are easily distinguished from hand-colored prints under 8-10x magnification by the presence of regular patterns (usually dots or pixels).  Since many printing processes do not involve dots, a lack of dots does not indicate handcoloring.  Under 8-10x magnification, a hand-colored print will show smooth planes of color marked by the irregularities (dark edges, for example) typical of watercolor paint.  Hand-colored prints will have spots where the paint goes over or under the black ink lines, places where the colors smudge or overlap, etc.

If you are having trouble determining whether or not your print is hand-colored, buy yourself a cheap loupe (typical cost $10 or less) from a photo supply store and examine your print with the loupe right on top of it.  It is generally not useful to use a loupe on a framed print because you cannot get close enough to the print to bring the image into focus.  Most of us know from first-hand experience what watercolor looks like.  Printed color usually shows a regular pattern of some sort (although it may not involve dots). If it doesn't have dots, but it doesn't look like watercolor, it is probably a different type of printed color.  Please also remember that there are quite a few hand-colored reproductions and facsimiles of Havells in circulation; hand-coloring alone is not proof of authenticity for an Audubon.

Return to TOP

Presence of a Plate Mark  

Plate marks are the rectangular ridge that completely surrounds the image and all of the writing in etchings and engravings.  Plate marks are sometimes trimmed away, or can be obscured, so they will not always be present or easily detected, but they appear on many of the prints we sell, and the presence of a plate mark is a valuable aid to authentication.  On Havells, plate marks on the largest of images are usually difficult to see (if they have not been trimmed away), but they are usually visible (and almost never trimmed away) on medium and small images.  Plate marks can also be found on the prints of many of the other artists we carry including Alexander Wilson, Prideaux John Selby, and Mark Catesby.  Plate marks are not found on stone or other types of lithographs, and so they are absent from works by Gould, Keulemans, and Brasher and all original Audubons except Havells. Catesby 2nd edition Hummingbird

Unfortunately with web photography it is not always easy to see a plate mark that would be easily discerned in person, but if you click on the photo at left, you should be able to see the medium-sized plate mark that completely surrounds the image of the Purple Martin (including the writing that is part of the image).

Havells were produced using copper plate etching.  A plate mark is formed by the pressure of the printing plate on the paper.  The plate leaves an appropriately  shaped indentation -- in the case of Havells, rectangular with rounded corners -- that completely surrounds the image and all text.  On rare occasions, small or medium prints are trimmed within the plate mark, but this is fairly unusual.  In cases where the images were  large relative to the sheet of paper, there may not be a visible plate mark, although you may still be able to feel the ridge along one or more edges of the print.  Although a plate mark is a helpful indicator of authenticity, the cautious collector must be aware of the fact that some facsimiles have "fake" plate marks (these can be identified as reproductions using other criteria).  Also, a small number of the original Audubon copper plates have been used to create restrikes, that is, prints produced in the 20th or 21st century by printers other than Havell. 

Return to TOP

Presence of a Watermark 

whatman watermarksThe watermark is best known of all authentication criteria; of the major Audubon editions, it applies only to Havells. The paper used for the Havell Edition was of two types, J WHATMAN and J WHATMAN TURKEY MILL.  Both types of paper had a watermark that can usually be seen when the print is backlit. At left is a composite photo (which can be clicked for a larger view) that shows examples of each of the two watermarks found on Havell Edition prints. The watermark is generally parallel to one of the LONG edges of the print. With J Whatman Turkey Mill, the watermark may be found with the year sitting adjacent to one of the long edges, whereas the J Whatman mark will usually be slightly more towards the center of the paper. Thus the watermark on a J Whatman sheet is less likely to be trimmed during the binding process, whereas there can conceivably be slight trimming of the year on a J Whatman Turkey Mill sheet if the binder was not careful to preserve as much of each sheet as possible.


The 18XX indicates the year of the paper's manufacture and will range between 1827 and 1838.


On rare occasions one can see the watermark without back lighting, but this does not happen often. Much more problematic, with very dark prints it can be difficult to spot the watermark, but it is always important to search for it. Lack of a watermark is strongly suggestive of a later printing (that is, a restrike of some sort). Lack of a watermark that theoretically ought to be there (that is, could not have been completely trimmed away) will create a serious problem for authenticating (and therefore selling) a print. Even if a print can pass authentication based on the other criteria, lack of a watermark (with no plausible explanation for its lack) may prevent the sale. Lack of a watermark (if there is a plausible explanation) may allow the sale, but it will certainly affect the price independent of the fact that the print is trimmed. The more adverse the impact of trimming on the watermark, the lower the market price of the print is likely to be.