A mezzotint star map

Matthew Edney

One problem with doing historical work is that there are some things you just have to take on faith. One great thing about doing historical work is the insight that what you took on faith is just plain wrong. I am not naming names, as such errors can be created by anyone working quickly, or who themselves take their predecessors’ statements on faith. I am sure that someone will point out my own failings in this regard!

Consider the white-on-black star charts by Christof Friedrich Goldbach, in his Neuester Himmels-Atlas of 1799. Each spread of this atlas contained two images. On the left, the stars in a portion of the sky:

Plate 7 left: area of the night sky with Boötes, etc.

Plate 7 left: area of the night sky with Boötes, etc.

On the right, the stars were joined by coordinate systems (with respect to both the celestial equator and the plane of the ecliptic) and the outlines of the figures of the constellations:

Plate 7 right: area of the night sky with the constellations of Quadrans Muralis, Boötes, Canes Venatici [ Jagdhunde ], and Coma Berenices [ Haupthaar der Berenice ]

Plate 7 right: area of the night sky with the constellations of Quadrans Muralis, Boötes, Canes Venatici [Jagdhunde], and Coma Berenices [Haupthaar der Berenice]

Now I read in a standard source that these images were printed from the same plates. The left-hand images were prepared and impressions taken, and then the extra information was added and further impressions taken. Made sense on the face of it. Also, the same source suggested that these were printed from copper plates but by a low-pressure relief method, such that the ink is lifted of the plate surface rather than from within the lines inscribed on the plates surface. Could be the case, so I took it all on good faith.

Until I had the occasion to look at this atlas in great detail, in preparation for a new exhibition at the Osher Map Library at University of Southern Maine.  And two things immediately leapt out at me:

1) each image has a plate mark, the small ridge formed when the paper was forced over the edge of a copper printing plate in a high-pressure, rolling press. This means that the maps could not have been printed from ink on the surface; in a rolling press, such loose ink would squirt everywhere! Upon careful study, it is clear that all of Goldbach’s plates are mezzotints. This technique was developed to mimic oil paintings: a copper plate is first covered in a very dense pattern of dots, so it prints as a solid, velvet black; then the holes are burnished and polished with something very much like a thimble, to produce shades of gray all the way to white.

2) each pair of images are actually from separate plates. Many of the right-hand plates are distinctly larger than the left-hand plates (each plate is prepared almost to the edge, so the edge of the ink is almost precisely the edge of the copper). And there are clear differences in the details of stars and things.

So, the nice, neat story first advanced is incorrect. The technique is not remarkable, but the effect of white stars on the black night sky was.

You can see the whole atlas at the Osher Map Library. (This link might break as OML’s website is about to be completely redone, so you can also try http://www.oshermaps.org/map/4406.0001.