Galileo's Moon Drawings

The purpose of this page is to bring together Galileo's surviving observations of the Moon, and the speculation regarding how and when they were made. An immense amount has been published on this subject, most of it inconsistent, and some of it obviously wrong. Please excuse the untidiness of our presentation while we, along with you, attempt to make some sense of it.

Frequent reference will be made to certain images published by and thought to have been drawn by Galileo. Particular emphasis is placed on a set of wash drawings appearing in a copy of the January 7, 1610 letter in which Galileo first announced his discovery of the moons accompanying Jupiter, on the engravings appearing in various editions of his book Sidereus Nuncius (first printed in Venice in mid-March 1610), on a sheet of wash drawings bound with Galileo's handwritten manuscript copy of Sidereus Nuncius at the National Library in Florence (BNCF), and on a brief series of perspective sketches of lunar features appearing on an undated sheet of calculations (also preserved at the BNCF). In the following we will show reduced size images of each of these with numbers assigned for identification. You may open the full-size original images on which these were based by clicking on the links in the following table. The selected image(s) should open in a new window from the host site (IMSS or BNCF). To the best of our knowledge, high resolution versions of the images in the recently discovered hand-illustrated possible printer's proof of Sidereus Nuncius are not yet available.

For the engravings published in Sidereus Nuncius the links will direct you to the copies appearing in the IMSS's electronic version of the National Edition of Galileo's works, which are presumably photographs of the versions appearing in one of the first copies printed in Venice. Because printing quality is never perfect, the appearance of these plates varies slightly from copy to copy; and greatly in later editions, such as those later published (probably without authorization) in Frankfurt and London. Several places where you can view other copies of Sidereus Nuncius over the internet are indicated on our Additional Information and Accessing Manuscripts pages. Several of the other versions are as good as the IMSS copies linked to above. They also show more clearly how the images were placed on the printed page, something that is not obvious from the electronic transcription at the IMSS.

Introduction

It is by now well known that Galileo was not the first to examine the Moon through a telescope. Human curiosity being what it is, the idea of turning the early telescopes towards the sky had occurred to many. For example, in his article on the invention of the telescope, Albert Van Helden reprinted selections from a French newsletter published in October 1608 mentioning a telescope recently demonstrated in the Hague. Not only was it said to be useful for military operations, but when pointed at the sky it revealed stars too dim to be seen by the unaided eye. It is also now well known that the English mathematician and adventurer Thomas Harriot produced primitive drawings of the Moon as seen through a telescope dated as early as July 1609; and others later claimed to have made astronomical observations during that year and before.

Although Galileo carefully recorded the date and time of his observations of Jupiter's moons and, later, of sunspots; few, if any, of his surviving lunar observations are dated. This is a little surprising since it must have occurred to him that things in that new uncharted world might change with time; yet it appears that he did not date them. There is no real scientific purpose for assigning dates to them now; yet it is a topic that has stimulated considerable interest and debate. Many believe that the Earth's Moon was the first celestial object Galileo observed, and hence the date of his first observation of it might pinpoint the start of the remarkable sequence of discoveries by this most famous of early astronomers.

The possibility of assigning dates to the lunar engravings printed in Sidereus Nuncius on the basis of their appearance and historical context seems to have first occurred to Guglielmo Righini, a noted solar astronomer, and longtime director of the Arcetri Observatory in Florence. Righini had close ties to Galileo, for not only is the Arcetri Observatory adjacent to Galileo's final home, but Righini's wife, Luisa Maria Bonelli Righini, was the longtime and much beloved director of the museum (the IMSS) where most of Galileo's artifacts have been collected. Righini came to the conclusion that what he believed to be the first of the four distinct engravings in Sidereus Nuncius must represent an observation made shortly after sunset on October 2, 1609. His finding was challenged almost immediately by Owen Gingerich, and later by Ewen Whitaker, a prominent selenologist (lunar astronomer) at the University of Arizona's prestigious Lunar and Planetary Laboratory. Whitaker, with equal confidence, assigned a date of November 30, 1609 (two months later than Righini's date) to the same engraving, which he also felt represented the first evening of Galileo's celestial observations.

Whitaker's work, perhaps because of his more substantial credentials in the field of lunar science, seems to have attracted more attention than Righini's, and has led to the rather romantic notion that Galileo, paintbrush in hand, first turned a newly made 20-power telescope on the night sky on the evening of November 30, 1609, instantly recognizing the topography of craters and mountains written in the ragged pattern of light falling on the crescent Moon. This is certainly the impression it gave to an anonymous reviewer of Whitaker's article writing in the March 1979 issue of the popular and widely-circulated astronomy magazine Sky and Telescope.

Lest the reader imagine that the anonymous reviewer had added something for dramatic flair, the following are Whitaker's own words from a 1984 review of Selenography in the Seventeenth Century in the Cambridge General History of Astronomy:

"[Galileo] first directed one of his home-made telescopes (the 20-power) towards a non-terrestrial object during the evening of 30 November 1609, the object being the four-day-old crescent Moon. The fact that he observed that body until almost moonset, describing the progress of sunrise on what has since been identified as a group of mountain crests in the Janssen area of the lower cusp, as well as making three separate sketches of the Moon, shows that he immediately grasped the importance of what was revealed."

Yet, it is fairly clear from Galileo's letters that he must have looked at, and understood, the Moon considerably before this. In an August 29, 1609 letter to his brother-in-law, Galileo mentions both his pleasure at receiving a lifetime professorship at the University of Padua in return for his recent demonstration of the telescope in Venice, and also his disappointment that such a position would mean he could visit Tuscany at most during the university recesses. In a January 30, 1610 to Belisario Vinta, acting as a secretary to the Tuscan court in Pisa, Galileo describes some of the many wonders to be announced in his forthcoming book (Sidereus Nuncius). According to the translation published by J. J. Fahie in 1903, Galileo notes that his observations of the Moon will be somewhat old news, since he "had already ascertained that the moon was a body very similar to the earth, and had shown our Serene Master, the Grand Duke, as much, but imperfectly, not then having such an excellent spy-glass as I now possess." Since Galileo was living in Padua, some 120 or more miles from Tuscany, this demonstration of the Moon to the Grand Duke evidently occurred on one of those rare visits alluded to in the August 1609 letter. Fahie guesses the visit took place in October 1609 and quoting we know not what document says that the Grand Duke "to his great surprise and delight was able to see that the moon was a body very similar to the earth." The January 30 letter demonstrates not only that Galileo was curious enough to look at the Moon with telescopes less perfect than the one used to obtain most of the observations reported in Sidereus Nuncius; but also, if the date of his demonstration to the Grand Duke can be determined with more certainty, that he was aware of the earth-like nature of the Moon considerably before the November 30 date suggested by Whitaker.

Gingerich and Albert Van Helden, are among the very few professional scholars who acknowledge that Galileo may have looked at the Moon and even made sketches pre-dating any of those that survive. But even they believe Whitaker correctly dated all the surviving drawings and engravings, except one. Yet it is hard to believe that everyone is so sanguine about the accuracy of Whitaker's dates, or even the possibility of dating lunar drawings. Such dating requires the exact identification of features portrayed in the drawings. Manchester University's Zdenĕk Kopal, a noted astronomer of Whitaker's era, and author/editor of several definitive 1960's books on the Moon, including the Pic du Midi Photographic Atlas, repeatedly expressed his opinion that the engravings in Sidereus Nuncius were purely schematic and qualitative, and that there was no definite correspondence between any of the features depicted and real features on the Moon. Unfortunately we do not know if Kopal was aware of Whitaker's 1978 article or ever expressed an opinion about it. Similarly, Righini published a very slightly expanded version of his own datings in a 1978 compilation of his works about Galileo; in which he mentions (in note F) being aware that Whitaker had a similar article in press; but we do not know Righini's opinion about Whitaker's paper. Moreover it appears that none of these early researchers has bothered to check very carefully the claims of the others. In particular, no one seems to have investigated the extent to which two different people could independently arrive at the same identifications for the features shown along the terminator; nor whether the Moon would have actually looked the way these experts claim on their stated dates.

As to the identification of features we have been unable to find a single instance in which Whitaker and Righini claimed that a object on one of Galileo's engravings represented the same feature on the real Moon. As to the prediction of what the Moon would have looked like from a particular place at a particular time, modern computer technology has made this, if not necessarily any more accurate, at least much easier than it used to be. This a luxury that was not available to Righini, Gingerich or Whitaker in the 1970's; and it has brought accurate predictions within the range of non-experts like ourselves. The man in the street can now look at a photograph (or make their own observations) under circumstances very close to those that should have existed on the days the experts claim Galileo's images were made, and ask themselves whether the resemblance of the features is enough closer at those times (as opposed to some other) to assign a definitive date. As non-experts, we find considerable reason to be skeptical about all the dates to Galileo's lunar drawings, with the possible exception of one showing a star emerging from occultation. In the case of Righini, we find no resemblance between Galileo's engravings and the way the features he claims to have identified would have looked on the dates he mentions. In the case of Whitaker, there sometimes seems to be a vague resemblance, but the number of features in Galileo's images that cannot be plausibly accounted for on those dates, and number of ones that would be expected to be prominent but cannot be found in Galileo's images so far exceeds the others as to make one wonder if the occasional resemblance might be caused by nothing more than chance. This is not to say that the dates assigned by the experts are necessarily wrong (if they are true images of the Moon at a particular moment then there are, after all, only a very limited number of possible dates on which they could have been based), but only that the means by which those dates were determined seem highly suspicious to us. We find nothing in the visual or historical record definitive enough to justify so supremely confident and unequivocal a conclusion as that of Whitaker quoted above. Indeed, the number of inconsistencies in the interpretation, and in the drawings themselves, are so great that we are led to suspect with Kopal and others that the surviving images are much more impressionistic than realistic. We would be happy to be proven wrong, but we suspect, for that reason, that attempts to match the images in detail to the actual appearance of the Moon at any particular moment will ultimately prove futile.

Whether exact dates can be assigned to a particular drawing or not, we find it extremely hard to believe that Galileo's understanding of the nature of the lunar topography sprung (as Whitaker claims) full-blown into his mind on a single evening. It seems far more likely to us that, like the anonymous observers mentioned in Van Helden's 1608 newsletter, Galileo would have had the curiosity to look at objects in the sky, including the Moon, even with his very first telescopes in the summer of 1609. The January 7 letter suggests that Galileo must have been able to make reasonably detailed observations of the Moon, and to have had some understanding of what he was seeing by at least December 11, 1609. But this understanding was probably the culmination of a program of observing that had started some time before that, in a process to which no precise starting date can be assigned.

Drawings in a Copy of Galileo's Letter of January 7, 1610

Galileo's most famous letter is undoubtedly that in which he announced his discovery of three tiny stars near Jupiter, which later proved to be moons orbiting the planet. The story of this letter has been told at some length by Galileo scholar Stillman Drake, who includes a translation in his 1976 article. According to Drake, the original, which once belonged to Galileo's son, Vincenzo, seems to have been lost. The most complete copy is in the Vatican Library, and it includes the copies of Galileo's drawings of the Moon reproduced below. A second copy, in Florence, lacks both the drawings and the final page. Ironically, this famous letter, apparently written in response to a request for information about his telescopic observations, may never have been sent. Drake regards Favaro's assignment (in the National Edition) of Antonio de Medici as the intended recipient as without basis (he favors Galileo's fellow mathematician/astronomer Enea Piccolomini). Drake believes that the second of the letter's two sheets, containing the announcement of the stars near Jupiter (not yet recognized as moons), the signature and date, was written some days after the lunar observations; and that instead of sending the letter Galileo used the second sheet to record additional observations of Jupiter on subsequent nights. These, according to Drake, he copied into a new journal, discarding the original page; although it seems difficult to understand how this scenario could be consistent with the existence of the Vatican copy, ending with Galileo's signature and the January 7th date. In any event, Drake believes there was a lapse of time between when Galileo completed his description of his lunar observations and when he signed and dated the second page. The quality of the original drawings, and exactly what they depicted can only be guessed at from the anonymous copies, and from Galileo's description of the originals in the text of the letter, which has been likened by many to a first draft of Sidereus Nuncius.

Because of their apparent crudeness, these little copies of Galileo's drawings seem to have attracted almost no attention from professional Galileo scholars. Drake dismisses them as being copies of such poor quality that they serve only to indicate "what diagrams were included with the letter." He does not bother to reproduce them in his article. Whitaker also does not discuss or reproduce these drawings because, he says, "none of these displays any recognizable surface features." Yet, for the most part, they correlate quite well with the verbal descriptions in the letter, and, to us, seem to have important implications as to the reliability of dates which have been proposed for the other extant drawings and engravings. This is because the lunar features described in the letter seem to be the same as those described in Sidereus Nuncius; yet the copies of the drawings from the letter suggest that at least some of those features may have been observed at lunar phases quite different from what one would gather from the published engravings.

To make it easier to discuss the individual drawings, we have adopted Whitaker's procedure of assigning a Letter-Number designation to each. The drawings we call L1-L5 are sketches of the complete Moon, while C1-C4 depict the play of light across a single crater.

L1: the Moon, four or five days after New. The boundary separating dark from light is rough, broken and confused. Light points protrude into the dark, and vice versa.
L2: [apparently an alternative version of L1] luminous peaks appear in the dark area, well separated from the light part. Over a period of some hours the luminous points grow and merge into the light area.
L3: shows the many dark spots visible in the lighted area. The dark spots have partially illuminated rims "facing the Sun". The spots are most frequent near the boundary between light and dark and, especially, in the lower [southern] horn of the crescent [the area we now call the Southern or Lunar Highlands]. In this area the spots are so numerous that they resemble "ice glass" [crackle ware]. The spots get smaller and fade at Full Moon, but return as the Moon wanes. Like a valley on Earth surrounded by mountains, the side of each spot towards the Sun is dark and the side towards the dark part of the Moon is light.
C1..C4: [Drake refers to these as "four schematic illustrations of different phases without any detail (and of smaller size than the other drawings)." We are unable to fathom his reason for believing they show the entire Moon. In the National Edition they are printed next to a sentence clearly stating they depict the changing pattern of illumination expected as the Sun shines at different angles on a particular crater with which Galileo was particularly impressed. Drake's own translation of that sentence is given in full below.]

The spots come in different shapes, usually very irregular. There is one of particular interest that is:
1. Near the middle of the Moon.
2. Perfectly circular.
3. Quite large compared to the others.
Galileo observed this crater both at the moment when the Sun began to light the far rim, and later as the Sun rose and began to illuminate the bottom. It gave him the impression of an amphitheater on Earth, or, more precisely, the province of Bohemia: a circular plain bounded by very high mountains. Galileo, referring to this Bohemia-like crater, describes drawings C1..C4 as follows [Drake's translation]:
"And its appearances before and after Full Moon are similar to these, noting that the shady part is always on the side of the Sun and the bright part opposite thereto, a positive indication that this is a perfectly round cavity surrounded by a high boundary."
[The drawings, apparently impenetrable to Drake, appear to us to show: in C1, with the Sun rising to the right, a thin sliver of light is first seen illuminating the extreme left edge of the crater. In C2 and C3, as the Sun continues to rise to the right the illuminated area spreads across the floor of the bowl from left to right. Finally, in C4, with the Sun setting to the left, the shadow of the crater's left rim has begun to spread across the floor, extinguishing the bright part.]
L4: the Moon near its quarter. In the south there is an immense dark gulf that digs into the lighted part. As the sun rises, within this gulf a triangular prominence spreads into the form of promontory. Later, other bright points emerge, initially isolated in the dark, but growing and merging with the lighted part, like the tops of mountains on Earth which are lighted by the rising Sun before the valleys to their east. Such effects are seen only in the lighter parts of the Moon; the large dark patches visible to the naked eye appear smooth and do not change in appearance as the angle of illumination changes, although here and there are little patches brighter than the rest of the area. The contrast between the light and dark parts of the Moon is similar to what Galileo believes would be the appearance of the rough continents versus the smooth seas of Earth as viewed from space.
L5: the Moon, five or six days old. One of the large dark naked eye spots has long, brightly illuminated prongs at its top and bottom. The prongs converge toward the east [i.e., to the left, opposite to the current IAU convention for east and west on the Moon], and extend into the dark side of the Moon as shown.

These drawings place a definite limit on the latest date at which Galileo could have begun his telescopic observations of the Moon, because in his comments on drawing L3 he mentions he knew that the crackleware effect in the Southern Highlands was hard to distinguish at Full Moon. In Padua, where Galileo was living at the time, the most recent Full Moon occurred on the morning of December 11, 1609, the New Moon was early on December 26, and at the time the letter was signed was not yet quite full again. With the exception of the little diagrams of the Sun rising and setting over a bowl-shaped depression, all the drawings appearing in the letter could have been made, judging from the phase depicted, between about December 29 and January 1 or 2. However, they could be older for Galileo must have been observing longer, since he indicates he had seen craters illuminated both from the east and the west. Such observations of the Sun rising and setting over craters in various areas of the Moon could have been made between December 11 and December 26. Finally, the illustration L1-L5 themselves indicate phases from a little after a New Moon to a little before a first quarter. From Padua, the New Moon immediately preceding the date of this letter occurred in the early hours of the morning of December 26, 1609. The drawings indicate phases that could have been observed between approximately December 29 and January 2. It is, therefore, conceivable that Galileo could have started his telescopic observations of the Moon as late as December 11, 1609 and still seen everything reported in the letter of January 7, 1610. However, as we mentioned previously, it seems likely to us that he would have started much earlier, even though the views through his earliest telescopes would have likely been very imperfect. Galileo may have heard about the telescope as early as May 1609, but he says he did not try constructing one until he received a letter about it from the French nobleman Jacques Badovere (the letter does not seem to have survived); hitting upon the method of constructing it a short time later; and demonstrating a working version of about nine power to various dignitaries in Venice between August 21 and 24 (see Rosen, 1951). Although Galileo might well have looked at the Moon through this instrument, its power was, by his own admission, insufficient to see most of the things reported in Sidereus Nuncius, including clear images of the lunar craters. On the basis of the preceding reasoning, the date on which Galileo began to correctly understand the nature of the lunar topography based on observations with his telescopes seems to fall after August 24 and before December 11, 1609; trying to place a more precise date on it than that is probably futile.

This letter (which is written in Italian) seems to be the source of the commonly held belief that Galileo began his astronomical observations with a sighting of the four to five day old crescent Moon. The letter actually says, by way of introduction to figure L1:

Et le apparenze da me nella luna osservate, sono queste.
Prima, cominciando a rimirarla 4 o 5 giorni dopo il novilunio, vedesi il confine ...
come nella figura apresso si vede.

And the appearances observed by me in the Moon are these.
First, commencing to look at it four or five days after New Moon, the boundary is seen ...
as seen in the adjoining figure.

It seems to us that Galileo has to start his description somewhere, and he may simply be saying that he will begin it with a drawing showing what one sees if one looks carefully at the Moon when it is 4 or 5 days past New. We are certainly not experts on Italian usage, but Drake's translation seems ambiguous enough to be read this way; particularly since, according to Drake, Galileo says he is showing how "the boundary is seen" rather than how "the boundary was seen". That is (if Drake's translation is correct), Galileo is presenting drawing L1 as a representation of a typical 4-5 day old Moon, not as a recollection of some particular one seen on a definite past date. He seems to be saying "If we commence our description with a careful examination of how the Moon looks when four or five days past Moon, we will see ..." Yet Drake unambiguously states his own belief that an observation of a five day old Moon on December 1, 1609 was Galileo's very first lunar observation, resulting in two "not very accurate" drawings.

Galileo was capable of writing in a direct and forceful way. If he had wanted to say "When I commenced my observations, the Moon was four or five days past New, and I saw the boundary as shown in the first figure ..." it would seem he could have easily done so. But he did not, either here or in Sidereus Nuncius (which appears to have been composed directly in Latin). Indeed, in Sidereus Nuncius Galileo used virtually the identical phrases to begin his detailed description of the Moon and introduce the first engraving, but chose to completely delete the words "first" and "commencing to examine," leaving no particular impression that this was his first observation:

The things I have seen by which I was enabled to draw this conclusion are as follows.
On the fourth or fifth days after New Moon, when the Moon is seen with brilliant horns, the boundary ...
as shown in the figure below.

The experts also allude to the passage towards the beginning of Sidereus Nuncius where Galileo says that after completing his 30 power telescope "forsaking terrestrial observations, I turned to celestial ones, and first I saw the moon from as near at hand as if it were scarcely two terrestrial radii away" (compared to what he believes is the normal 60 radii). This would seem to imply that viewing the Moon through a 30 power telescope was his first celestial observation. However, not only does he not mention what phase the Moon had, but this is clearly an unreliable description of his first celestial observations, since we know from the January 7th letter that he had already thoroughly observed the Moon before ever completing the 30-power instrument, for he says (in the letter) he is "about to finish" making it. So the passage may merely mean that the Moon was the first celestial object on which he tested his new 30-power telescope and he was gratified by how near it looked; not that this was his first celestial observation ever.

To emphasize that things may not be as tidy as many seem to think, it should be noted that in the January 7 letter Galileo most definitely describes a vast gulf with a triangular peak as having been observed in connection with drawing L4, in which the Moon is both shown and said to be near its first quarter. Yet in Sidereus Nuncius Galileo attaches virtually the identical words to a gulf and triangular peak illustrated in his first engraving: of a four to five day old crescent Moon (see engraving E1 below). The only thing lacking here is the statement that the merging of the promontory with the lighted portion of the Moon became complete just as the Moon has about to set. In dating engraving E1 to November 30, 1609 (at 8 pm Padua time), Ewen Whitaker attaches considerable significance to this triangular peak and the three surrounding points of light, all of which he claims to have conclusively identified. But if Galileo was drawing a feature he had seen not on a four day old crescent Moon, but rather much closer to the first quarter, as he says in this letter, then Whitaker's identifications, and hence his date and time, would be incorrect.

Engravings Appearing in Sidereus Nuncius

Galileo's best known images of the Moon are five engravings (one of which is a duplicate), that appear in his book Sidereus Nuncius, published in mid-March 1610. The four unique engravings are shown at left. The red numbers E1-E4 are those assigned by Ewen Whitaker. The dates assigned to these engravings by Whitaker, Righini, and others, are given in a table later on this page.

Note: the reproductions shown here are photographic prints from the National Edition of Galileo's works. At about the 7 o'clock position on E2, E3 and E4 is what looks like the letters "V.T.". These are presumably the initials of someone who worked on the negatives. These letters do not appear in the printed copies of Sidereus Nuncius.

The two things that have most attracted modern readers' attention are the identification of the large crater shown on the terminator in engravings E2 and E4, and the date on which engraving E1 was made.

The special interest in engraving E1 is because many read the text of Sidereus Nuncius to imply that a sighting of the four to five day old Moon was the first of Galileo's astronomical observations, and that the other lunar drawings followed in sequence. As we have indicated in the previous section, this interpretation is by no means obvious to us and seems to have very little basis in the text. Galileo may have felt that if one begins with the New Moon, the first phase in which the unusual, jagged nature of the terminator is strikingly obvious is when it reaches this age; and that he could go on from there to describe how the appearance of the Moon changes during the remainder of the lunar cycle. In other words, he may have chosen the order of his descriptions for literary effect and clarity of exposition, rather than as a literal transcription of the sequence of his observations.

In the text it is also not always clear which words go with a particular illustration, and a number of comments are repeated many times. We believe, however, that the points Galileo wishes to make about these four engravings can be summarized as follows [again we try to repeat his words as accurately as possible and place explanatory comments in square brackets]:

Galileo's description of the triangular peak illustrated in engraving E1 is so detailed and exact, that one would think it could be easily matched to some easily identifiable set of features on the Moon. However it is not totally clear if at the point illustrated by E1 the triangular peak is the large blob shown off the terminator, towards the bottom; or the sharply pointed promontory above it; or if he is possibly illustrating both facets at the same time. Righini was quite confident that Galileo was describing the Sun rising over the craters known as Baco, Breislak, Clairaut and Cuvier on the evening of October 2, 1609. Whitaker was equally confident that Galileo's triangular peak is "the illuminated centre of Janssen floor plus Fabricius west wall" as it would have appeared on the evening of November 30, 1609; and he identifies the three points of light as (1) the southwest wall of Janssen, (2) the west wall of Brenner, and (3) the west wall of Metius. Neither Righini nor Whitaker produced photographs of their proposed regions which look convincingly like Galileo's engraving, let alone photographs taken under the lighting conditions expected to prevail on the evenings they propose. In a later section of this page we will address the question of whether the features mentioned by Righini and/or Whitaker would actually have looked as they do in Galileo's engraving on the dates and times they suggest.

A more pressing question is whether Galileo actually observed this sequence of events along the terminator of a four to five day old Moon, as Righini and Whitaker assume; or if he has, for literary effect, interpolated into that discussion (and illustration) something he actually observed much closer to the first quarter. The reasons for thinking Galileo may have done this are first that he describes what appears to be the exact same sequence of events in his January 7, 1610 letter, and associates them there with drawing L4, which shows a completely different phase from E1, much closer to first quarter. Second, although Galileo introduces engraving E1 in Sidereus Nuncius as representing a four or five day old Moon, he begins the paragraph about the triangular peak illustrated in E1 with the statement that he is about to describe some striking events that he "observed when the moon as approaching first quarter". Whether Galileo regarded a four to five day old Moon as "approaching first quarter" is anybody's guess. If Galileo is indeed showing a composite illustration giving an impression of the appearance of the four day old terminator together with a detail of events occurring at the first quarter, and possibly even illustrating two different stages of the peak merging with the terminator in the same drawing, then dating that drawing by the methods of Righini and Whitaker (or any other method) is quite impossible since the engraving would represent no single date.

We are not alone in thinking it possible that the engravings shown in Sidereus Nuncius should not be taken entirely literally. Gingerich and Van Helden believe, for example, that Galileo produced the engraving of the first quarter Moon, shown as E2, above, by manipulating the watercolor drawing labeled F5 in the following section; copying the pattern of light and dark seen along the terminator at a rather different phase. Kopal, of course, is almost notorious for his claims that Galileo's drawings are only schematic impressions of the nature of the lunar surface.

As to the large crater shown on the terminator in engravings E2 and E4, Righini believes this is not a single crater, but an amalgam of the craters Purbach, Regiomontanus, Werner, Blanchinus and La Caille -- a region, he says, that might, at low resolution, look as Galileo has depicted it. Whitaker followed Gingerich in identifying the large crater as Albategnius. Kopal once hazarded a guess that it might be Ptolemaeus. Many other opinions have been expressed as well.

Inferior copies of these engravings, sometimes even printed upside down, appear in pirated copies of Sidereus Nuncius published in Germany (see the web exhibit from the Linda Hall Library), and were later unwittingly reproduced by others. To add to the confusion, Galileo's Venetian publisher, Thomas Baglioni, reprinted the four distinct plates from Sidereus Nuncius and inserted them in some copies of Giulio Cesare Lagalla's 1612 book De phoenomenis in orbe Lunae (van de Vyver, 1971), causing some people to mistakenly attribute the observations to Lagalla.

In his 1998 book about the art of the printed book, Adrian Johns shows the complete set of lunar images printed in Venice (1610), Frankfurt (1610), London (1653) and London (1658) reproduced from copies of those editions in the archives of the California Institute of Technology. It seems unfortunate that the Frankfurt printers produced rather crude copies of Galileo's original plates, for it is evident from Johns' presentation that the London engravers, in 1653, copied the Frankfurt plates so carefully that it is almost impossible to detect any difference. Nonetheless, a comparison of the images in the various editions give us a realistic sense of how accurately a 17th century engraving can be expected to represent the artwork on which it was based.

In the following we compare engraving E1 as it first appeared in Venice in 1610, in the second volume of an early version of Galileo's collected works (the Le Opere printed in Bologna in 1656), and in two editions of Sidereus Nuncius printed in London.

The London plates of 1653, as just mentioned, are virtually identical to the Frankfurt plates of 1610 (not shown here). The Bologna plates (we have only this very low resolution digitization from the IMSS Digital Library) appear to have been copied directly from the Venice printing. What strikes us is that although the various engravers differ considerably in their approach to representing the pattern of light and dark on the illuminated side of the Moon, they have all rather faithfully copied the shape of the terminator and carefully reproduced the points of light shown beyond the terminator in the dark part, both in number and placement. The appearance of the illuminated side was probably confusing and ill-understood by the engravers, but the stark black and white of terminator seems to have been easy to copy. Our conclusion is that the 17th century engravers were capable of very exact reproduction when they wanted to (as in the London plates versus the Frankfurt ones from which they were copied); and that even when they simplified shapes in the lighted part (as in the Frankfurt and Bologna plates versus the Venice originals from which they were copied), they retained with considerable accuracy the shape and pattern of bright dots along the terminator. It would seem to follow from this that in preparing the Venice plates, the engraver is likely to have quite accurately represented the lunar terminator as it was depicted in the original artwork given to him by Galileo.

Cherrington's Inversion Theory

Engraving E1 (Venice, 1610)	Engraving E1 (Bologna, 1656)	Engraving E1 (London, 1653)	Engraving E1 (London, 1683)

from Gallica IFN-2600224	from IMSS BD 300948_2	from EEBO Microfilm G291A	from EEBO Microfilm G293

In preparing the introduction to his popular 1968 book, Ernest Cherrington, a person very well experienced in observing the Moon with small aperture instruments, consulted a copy of Institutio Astronomica by Pierre Gassendi in the rare book collection of the Lick Observatory. This book, published in 1653, contained a reprint of Sidereus Nuncius with plates that Cherrington assumed faithfully reproduced the originals, and he included photographs of them in his own book. The book Cherrington refers to seems not to be the original of Gassendi's work, but rather a compilation put out by Jacobi Flesher in London. As well as the books by Gassendi and Galileo, the compilation includes the astronomical introduction to Kepler's Dioptrice, all in the original Latin. Four different copies of the 1653 printing (Wing G167, G167A, G291 and G291A - the last mislabeled in the catalog as being from 1643), may be viewed at Early English Books Online, along with a re-issue put out by Henry Dickinson in 1683 (Wing G293). The plates used for the lunar illustrations in the 1653 Sidereus Nuncius reprint are, in fact, extremely accurate copies, not of Galileo's originals, but rather of the degraded plates in the Frankfurt edition, and some were similarly printed upside down.

Cherrington correctly recognized that the version of E4 he was seeing in Flesher's reprint was a last quarter Moon printed upside down. Cherrington also decided that engraving E1 (printed right side up), was also upside down. After turning it around, Cherrington decided it represented a waning crescent, about 26 days old (that is, a view few days before the New Moon). He apparently thought he recognized the promontory around the crater Gassendi near the Moon's western limb, which does indeed, for a brief time around the 26th day, jut into the dark portion of the Moon much like the large pointed promontory Galileo shows above his triangular peak. However, this interpretation is quite inconsistent with Galileo's very clear statement that the engraving illustrates what he saw four or five days after the New Moon, and includes a depiction of sunrise over the triangular peak (as Cherrington would surely know, one sees only sunrise events as the Moon is waxing and only sunset events as it is waning). Cherrington also believes that engravings E2 and E3 (printed the same way they appear in "correct" copies of Sidereus Nuncius), are mirror images, inverted left-right relative to the view actually seen through a Galilean telescope. As he points out, such printing errors are not uncommon, even in modern times. His reason for suggesting this is that he believes the large basin near the top of E2 (with bright crab-link pincers extending into the dark side) to be Mare Imbrium, a huge lunar feature, readily visible to the naked eye, that is always to the left of center in the normal view. According to this theory, E2 and E4 are two different pictures of the same phase: an approximately 23 day old last quarter Moon.

As can be seen at left, given the information available to Cherrington, his inversion theory is not nearly as off-the-wall as it might seem. Viewed in this way, the plates do look remarkably similar, and the richly cratered region, now to the left of center, can be plausibly identified with the Moon's southern highlands. However, this seems to be the area Galileo refers to as the "peacock's tail" in his text, and mentions that as a sight visible as the Moon is waxing. Also, if Cherrington had had access to the original Venice engravings he would have noticed that the pattern of dark smudges in the upper left of the inverted view, only vaguely suggested in the 1653 reproductions, are, in the originals, unmistakably meant to represent the pattern of large lunar mare in the Moon's eastern hemisphere (that is, they should be on the right side as viewed from the northern hemisphere). So, although it is quite possible that the original engraving might have been faked in some way from an observation of the Last Quarter Moon, it was certainly meant to illustrate the First Quarter.

Our previous comments about the accuracy of the 17th century engravers notwithstanding, it is interesting to note that the engraver of the Frankfurt edition, apparently thinking such details unimportant, chose to show the shading inside the large crater at a 45° angle, rather than with a vertical boundary, parallel to the terminator, as had been the case in the original plates. All the later engravers followed his lead, copying the Frankfurt engraver quite exactly. On the real Moon, the boundary is, of course, vertical.

Engraving E4 (London, 1653)	E2 (1653 inverted)

from EEBO Microfilm G291	from EEBO Microfilm G291A

Cherrington believes the large crater on the terminator is Deslandres, which as we indicate on our homepage, is an attractive candidate as to size and position, but cannot be reconciled with Galileo's description of its floor as being bright on the side away from the Sun, and dark on the side towards it. Cherrington believes Mare Imbrium is also illustrated at a nearly opposite phase in E3 (which, as indicated, he also thinks is inverted left-right), with the shadowed sides of the large craters Clavius, Maginus and Plato causing indentations along the limb. To reconcile this interpretation of E3 with Galileo's clear description of this plate as being a representation of the Moon "before the second quarter," Cherrington suggests that Galileo means quarters of the lunar cycle, rather than the phases of illumination known to nearly everyone as first and last (or first and second) quarter; so that "before the second quarter" (of the cycle) actually means before the first quarter phase. This tortured explanation does not square with Galileo's use of the same terms (first and second quarters) in his discussion of the ashen light later in Sidereus Nuncius. So we find it hard to believe that E2 and E4 are really two different images of the last quarter Moon, one accidentally misprinted, making it look like the first quarter. Cherrington does not mention the large dark zones shown in E2 on the illuminated side of the Moon, which, even in the inferior copy in Flesher's reprint, resemble the naked eye markings seen on the east side of the Moon at first quarter, and have no obvious resemblance to the markings seen on the west side of the Moon at last quarter.

In fairness to Cherrington, it should be noted that it is only a guess that the lunar engravings were printed the way Galileo intended in the first edition of his book. As we observe on our Photo-Drawing Comparison page, his illustrations of star fields were printed in different orientations even in the "good" editions issued from Venice in 1610, and some are rotated both relative to the way he depicted them in his notebooks and to the way they would normally appear to an observer in the Northern Hemisphere. The printing of the lunar plates was similarly inconsistent. For example, the view of the first quarter Moon in the 1610 Venice edition at the National Library of France (see Illustration set IFN-2600224), reproduced on their Gallica website, is upside down. It has also never been obvious that the final lunar illustration in Sidereus Nuncius, where the printer has repeated plate E4, is the illustration Galileo intended. From the text one has the impression he was trying to illustrate something other than what was shown the first time E4 appears, and hence intended the printer to shown some other view.

The desire (and ability) to recognize familiar patterns in what we see is innately human. We describe Cherrington's unconventional views at such length only because they demonstrate the extent to which one can see what one wants to see. Somewhat like looking at an ink-blot test, Righini, Whitaker and others who have spent as much or more time than Cherrington puzzling over Galileo's engravings come to recognize different familiar patterns and thereby attach completely different interpretations to the same things. It is very difficult to know who, if anyone, is correct.

A Possible Printer's Proof of Sidereus Nuncius

On March 28, 2007, the University of Padua announced, with considerable fanfare, what they claim is the discovery of a previously unknown copy of Sidereus Nuncius. Simultaneous announcements were released in Italian, English, and Spanish-language newspapers, as well as on Italian language television. A German-language version (with a higher-resolution scan of page 9b) was also posted two days later on Spiegel Online. According to the releases, the book belongs to New York art dealer Richard Lan, who claims he obtained it from a source he refuses to identify in South America. Apparently this copy is in all ways identical to the others except that in the places where the five lunar engravings are normally found, one finds instead hand painted watercolors, which resemble the well-known watercolors described in the next section. At the instigation of Lan, the book was examined by a team of experts lead by historians William Shea and Horst Bredekamp. The team supposedly determined not only that this unique volume was a kind of proof copy used in the preparation of Sidereus Nuncius, but that the paintings it contains were executed by Galileo himself (whose signature apparently appears on the title page). This find has been much promoted by Bredekamp, who apparently features it in a new book of his own about Galileo's artistic output (Bredekamp, 2007). The release of this book (which we have not seen) triggered a second wave of news stories, the most accessible of which (to English-speaking readers) is probably an August, 2007 article that appeared in Time magazine (Israely, 2007). The Time magazine article implies that Bredekamp believes these are original images recorded by Galileo (working at the eyepiece) directly onto the pages of the proof copy.

The following figure brings together the pages in the "printer's proof" that show the five drawings (which, for convenience, have been labeled P1-P4b) as displayed in the University of Padua press releases and the Time magazine article, and places those drawings next to the engravings (E1-E4b) that appear on the same pages in the normal copies of Sidereus Nuncius. Note that the pages in Sidereus Nuncius are printed on both sides, but only the front sides are numbered. The notation "Page 9b" means that engraving E2 is printed on the back of page 9. In normal editions of Sidereus Nuncius, page 10b (the back of page 10) contains two lunar images but no text. One of these images is a repeat of the engraving printed on the numbered page 10 (the opposite side of the same page). Whitaker called this repeated engraving E4. Since the "printer's proof" contains different watercolors at these two positions, we use the notation E4a/E4b and P4a/P4b to distinguish the versions appearing on page 10 and 10b.

In images where there is printed text (pages 9, 9b and 10), the scale has been adjusted to make the layout of the printed text match in the normal and hand-illustrated versions. Relative to the type, it appears that the diameter of the watercolors on pages 9 and 9b is slightly larger than that of the engravings.

**Media Releases of "Proof" Paintings Compared to Published Engravings**

E1-E4b © U. of Oklahoma; P1, P2 © Corriere della Sera ; P4a © Richard Lan/Time Inc. ; P3, P4b © TGR

Since we know of Bredekamp's comments only as filtered through the news stories, our impressions of them are probably somewhat distorted. The notion that the drawings in this book could be "live" images recorded at the eyepiece seems especially fanciful to us, particularly since according to that scenario the words describing the observations would have had to have been committed to type before the observations were made. We don't know enough about the printing practices of the day to make any intelligent comment on the probability of this being a "printer's proof"; but if it were, the extra effort required to duplicate the drawing for engraving E4 in two places (P4a and P4b) seems odd, since a simple notation "same as drawing on previous page" would have seemed sufficient. Also, given the highly corrected state of the BNCF manuscript copy of Sidereus Nuncius it seems difficult to believe a working "printer's proof" of the same work would not be marked with last minute handwritten notations, corrections and instructions.

Another possibility, that does not seem to be mentioned in the press releases, is that this book might be some kind of presentation copy (rather than a proof) in which hand drawn illustrations were substituted for the normal engravings as a kind of special favor to the recipient. This might account for the apparently pristine state of the pages; but such an interpretation is plausible only if this was a known practice of time. Should this be the case, the illustrations would not be originals, but rather copies of whatever source material had been used for preparing the engravings, and such copies would not necessarily be the work of Galileo himself.

Watercolor Drawings Bound in the Manuscript Copy of Sidereus Nuncius

The National Library in Florence (BNCF) owns a heavily-corrected manuscript copy, in Galileo's handwriting, of the complete text of Sidereus Nuncius, referred to by Antonio Favaro in his introduction to Volume 3 of the National Edition of Galileo's works, as the first draft (there is apparently another manuscript in Germany; we do not know if it contains drawings). This document is bound in a folio along with other documents, including fragments of two other manuscript copies of Sidereus Nuncius, one apparently an Italian language translation by Galileo's assistant Viviani. The complete manuscript ends with a blank page. Sandwiched between that and the first page of Galileo's journal of Jupiter observations are the following seven drawings of the Moon, executed in either watercolor or sepia ink.

In 1975, Owen Gingerich described the paintings as having been executed in brown ink wash on a single "stiff folded piece of drawing paper". However, a 2003 article by Gingerich and Albert Van Helden suggests that it was Favaro himself who assembled these documents (something not entirely obvious from Favaro's own introduction), and says the drawings appear on "two single sheets of watercolour paper." We understand that the 2003 description of these as separate pages is the correct one (Gingerich, 2008, private communication). That the two separate pages could be confused for a single folded sheet suggests that the paper (which Gingerich says is very different from the other pages in the manuscript) must be quite similar -- which in turn suggests that the drawings came from the same source.

Whoever assembled the folio numbered the front side of each page in the folio consecutively, without respect to the particular manuscript involved. The two pages containing the lunar drawings were bound so that the drawings faced outwards (away from each other). This places them on the front side of the page numbered "28" and the back side of the page numbered "29". The following illustration shows the two sides of the two pages; but to juxtapose the lunar drawings, the back side of each is shown before the front side. In reality, when one looks at the front side of page 28, one sees the front side of page 29 through the torn corner in the lower right. For purposes of identification, the red F1..F7 shown next to each painting is the designation assigned to that drawing by Whitaker in his 1978 article.

Folio Page 28	Folio Page 29

Adapted from original images © BNCF

On page 29, the rectangular diagrams with notes are horoscopes, one complete, the other incomplete. Owen Gingerich seems to have been the first to determine that the complete one corresponds to the date of May 2, 1590. Guglielmo Righini, who did not mention these drawings in his paper about dating the engravings in Sidereus Nuncius, later identified May 2, 1590 with the birth date of Cosimo de' Medici. Since Sidereus Nuncius was dedicated to Cosimo, this seems to strengthen the association of this pages with that book; however, the lunar drawing (F7) that accompanies the horoscopes appears unrelated to the book (it is neither described nor pictured there). We are also not entirely sure whether these horoscopes are thought to be in Galileo's handwriting on not. In the National Edition, Favaro chose to edit out the horoscopes. Facsimiles of these pages (printed sideways in the print edition and upside down in the on-line IMSS version) appear after the text of the manuscript (as in the BNCF folio), after which Favaro has added facsimiles of two pages containing manuscript drawings of stars in the constellation Orion.

Our understanding of Latin and Italian is far too limited to understand the documentary evidence (if any) that links these drawings to Galileo and/or the composition of Sidereus Nuncius. Although there is a certain resemblance between the drawings and the engravings, in no case do the features shown along the terminator of a drawing exactly match the features shown in any of the engravings. Since there is a certain resemblance between all images of the Moon at a particular phase, we find it entirely possible that these drawings could have been made at a completely different time, and possibly by someone else. How and when they came to be inserted as pages 28 and 29 in the BNCF manuscript, or what might be the significance of the numbers appearing near the Moon's north pole in the drawings Whitaker has numbered F1, F4, F5, F6, and F7 are both mysteries to us. We have no idea why drawings F2 and F3 are unnumbered (Whitaker says F3 is numbered "2", but we cannot see this -- perhaps he was looking at the "28" in the upper right corner, which is the page number in the manuscript folio), or why the numbers "2", "3", "6" and "7" were not used. The numbers that were used seem to have been assigned roughly according to the sequence of the lunar cycle, but "5" (Whitaker's F5) is clearly an earlier phase than "4" (Whitaker's F6).

The best evidence we can adduce connecting these drawings with Galileo is the star shown to the left of the dark edge of the Moon in the drawing Whitaker labeled F7 (and someone else numbered 8). Although it is quite possible that someone simply put the mark there to indicate that this was a particularly fine illustration of the ashen light on the dark side of the Moon, Owen Gingerich suggested this might represent a conjunction of the Moon with a bright star, although he was unable to locate such an event. Ewen Whitaker later noticed that the Moon, as seen from Padua, occulted a 4th magnitude star in Libra on the morning on January 19, 1609. The star would have emerged from the dark limb shortly before sunrise at close to the position indicated, and the phase of the Moon would also have been extremely close to that shown. Occultations of 4th magnitude stars are not extremely rare (they would be expected, on average, perhaps once a month), but the coincidence of position angle and phase of the Moon is extremely unusual, and suggests that Whitaker is indeed correct regarding the date of this image. Since it would appear that Galileo did not share his astronomical observations with anyone before the publication of Sidereus Nuncius in March (Drake believes that to avoid being preempted by others, Galileo chose not to mail the January 1610 letter after discovering the moons of Jupiter), if this sketch was indeed made on January 19, 1609, then it would seem likely that the original of this image was drawn by Galileo. However, since there is apparently nothing to link this drawing to the ones on the other page (other than the similarity in paper and style), it is impossible to conclude anything, from this evidence alone, as to whether they are originals or copies and whether they date from the same or different periods.

It should be kept in mind that in preparing Sidereus Nuncius, Galileo had nothing to go by but his memory and whatever images he had managed to commit to paper. In 1975, Gingerich declared that these were careful studio drawings adapted from an archive of original sketches, since lost, which had been made at the telescope; an opinion with which Whitaker strongly agreed. By 2003 Gingerich had changed his mind, and declared, along with Van Helden, that not only are these the original sketches prepared by Galileo while looking through the eyepiece of his telescope, but that they are the ONLY sketches that were available to Galileo during the composition of Sidereus Nuncius. As a result, each of the four unique engravings appearing in Sidereus Nuncius must have been based on one of these drawings. In particular, since, according to this hypothesis, Galileo lacked a drawing of the first quarter moon, the engraving of that phase which appears in Sidereus Nuncius was achieved by distorting drawing F4.

Gingerich's Distortion Theory

In his 1975 comments on Righini's pioneering dating effort, Owen Gingerich noted what seemed to him a striking resemblance between engraving E2 and one of the wash drawings (F4) to be explained in the next section.

Gingerich's comments are worth repeating for the extent to which they demonstrate the highly subjective nature of attempts to identify features in Galileo's drawings and engravings. He says:

Engraving E2	Drawing F4

adapted from image © IMSS	adapted from image © BNCF

"... although there is no manuscript drawing of the moon at first quarter, the drawing labelled "3" [F4, below] made a few days earlier than the first quarter yields a very close correlation between features on the terminator and the one purporting to be the first quarter in Sidereus Nuncius [engraving E2]. Furthermore, the dark maria are reasonably represented on the manuscript drawing, but bear only a grotesque approximation to reality in the printed drawing. I would like to suggest that the printed drawing is a highly distorted and derivative version or the manuscript drawing."

Although we would have to agree that engraving E2 may well have been derived from F4, or a drawing like it (for the sea shown straddling the terminator near the top of the E2 is never seen in that position); our reaction on comparing the drawing with the engraving is almost exactly the opposite of Gingerich's. We see no obvious correlation between the features shown along the terminator (the two terminators seem at least as different as in the comparison of E2 to E4, inverted, shown earlier) and the maria seem to us much more accurately represented in the engraving than in the drawing. In fact to represent the maria in the northeast quadrant as accurately as he did (see the corresponding image through the Galilean telescope on our home page), the engraver must have had some input other than vague blobs in the drawing. They are admittedly not perfect in the engraving, but they seem a lot closer to reality to us.

Hevelius (1644)	Predicted Terminator

adapted from image © Cambridge University Press

The possibility that Galileo may have instructed the engraver to "improve" his drawing by straightening the terminator is certainly not to be dismissed. Even the later Polish astronomer Johannes Hevelius, who prided himself so much on the accuracy of his work (which he felt was much superior to Galileo's) that he made his own engravings, was not above doing this. In his Selenographia he printed at least one image in which the terminator appears to have been shifted for artistic effect. The example shown here, which appears on page 54 of Whitaker's 1999 book, is carefully labeled to indicate that it represents the 7 day old as seen from Gdansk at 7 hours after the meridian on March 15, 1644. According to a modern ephemeris, the Sun, as viewed from Gdansk, crossed the meridian at 10:55 UT, so this observation was, according to Hevelius made at about 17:55 UT (of course the drawing must have taken some time to complete, and he may be referring to mean solar time, which would be a few minutes different, rather than the literal meridian crossing; but this time should be close enough for the present purpose). A prediction of the position of the lunar terminator at that hour, based on the JPL lunar ephemeris, agrees very well with Hevelius' depiction of the features along it, including the prominent crater Albategnius, which he shows as the dark circle just below center. However, although the relation of the terminator to the features agrees with the engraving, it appears that the terminator should be significantly bowed to the left -- and not the perfectly straight line that Hevelius chose to show. Now, it is possible (although extremely unlikely at this scale) that the librations assumed in the JPL ephemeris do not extrapolate back properly to the 17th century -- that is, Albategnius might not have been precisely where JPL thinks it was relative to the Moon's limb; but the predicted shape of the terminator at 17:55 UT depends only on the relative positions of the Sun, Moon, and Earth, and those could not possibly be enough in error to account for the difference between the prediction and Hevelius' representation of it. Although the fudging is slight, it seems unlikely that Hevelius could have actually seen a straight terminator on this day: the exact moment of the terminator being a straight line would have occurred in the morning, some nine hours earlier, at around 9:00 UT, just as the Moon was rising with the Sun high in the sky. Hevelius, like Galileo, appears to have wanted a perfect First Quarter Moon for his book; and not having a perfect observation, he adjusted whatever he had.

In the captions to his figures, Gingerich invites us not only to compare engraving E2 to drawing F4, but also to make several other comparisons. Unfortunately, these directions are not elaborated in the text, so it is very difficult to know what he hopes for us to see.

Drawing F5	Engraving E3	Engraving E4

adapted from image © BNCF	adapted from image © IMSS	adapted from image © IMSS

In looking at engraving E3, Gingerich's caption instructs us to "note the different placement of the bright points at the terminator compared to the drawing [F5]." We would certainly agree that the bright points along the terminator are different in these two images; but we find it puzzling that in comparing E2 to F4 we were told to note "the very close correlation between" the points. They look equally different to us in the two examples. Finally, Gingerich tells us to compare E4 to E3. We have no idea what we are supposed to see; nor why he doesn't mention drawing F6, which looks at least as much like engraving E4 as drawing F5 looks like E3.

It is of interest to note that Ewen Whitaker would later declare that drawing F5 and engraving E3 were made at the same hour, implying that to him the points shown along the terminator represent identical features, even though they looked different to Gingerich. In this case we would have to agree with Gingerich: they look quite different to us. Based on our earlier comparison of the engravings copied from one printing of Sidereus Nuncius to the next, we find it very hard to believe that a 17th century engraver given F5 as his artwork would produce E3. On the other hand, in comparing drawing F4 to engraving E2, Whitaker decided they were based on observations 24 hours apart, implying that to him there was no correlation between the points along the terminator in the two images, even though Gingerich had declared them closely correlated. In this case we have to agree with Whitaker: we see no correlation.

Whitaker's Sequencing of the Drawings

Gingerich and Van Helden's view that Galileo had only his seven wash drawings to go by in preparing the engravings for Sidereus Nuncius was evidently not shared by Ewen Whitaker, who assigned a unique day and hour to each drawing and engraving. In only one case does the date assigned to an engraving (Whitaker's E3) exactly correspond to the date of a drawing (F5). Whitaker's belief is that the other three engravings were based on drawings since lost, perhaps by the printer.

As to the surviving drawings, Whitaker asserts with great confidence that the four he labeled F1..F4 were copied in the studio based on observations made at the following dates and times in 1609:

This sequence is based solely on Whitaker's claimed ability to recognize features along the terminator. Now Whitaker is highly respected selenographer who has made many commendable contributions to his field based on detailed analysis of photos which he can undoubtedly date with great accuracy, but there seem to be some grounds for wondering whether he really has this ability when applied to Galileo's drawings. For example, to the man in the street, it would appear that the position of the terminator in drawing F2 (that is, its distance from the right-hand limb) is about mid-way between that shown in F1 and F3, if anything closer to its position in F3. Yet Whitaker would have us believe that the interval from F1 to F2 represents the change in just 2 hours, while that from F2 to F3 is 24 hours. Recognizing this problem, Whitaker came up with his "qualitatively correct but geometrically inaccurate" drawings. That is, he believes Galileo drew certain features with great accuracy but placed them in the wrong positions.

F1 : Nov. 30; 15:00 UT	F2 : Nov. 30; 17:00 UT	F3 : Dec. 1; 16:30 UT	F4 : Dec. 2; 16:00 UT

adapted from image © BNCF	adapted from image © BNCF	adapted from image © BNCF	adapted from image © BNCF

There is undoubtedly many an amateur artist to whom this theory will have great appeal, for they have themselves created drawings in which the various elements were incorrectly placed relative to one another. Indeed, nearly all 17th century astronomical drawings we have seen suffer from geometric inaccuracies. However there is another problem. Allowing that the difference in position of the terminator in F1 versus F2 might be an artifact of geometric distortion, we are unable to understand how a "qualitatively correct" pattern of dots on the Moon's surface could have changed so totally in two hours: dark craters that were present in F1 have disappeared in F2 while new ones have appeared; and the terminator is completely different. Now a few features on the Moon can change significantly in two hours, but we are unaware of such a total transformation, particularly in the appearance of the illuminated crescent. Once again to the man in the street, it would seem that if the drawings show a total transformation when the actual change is slight, then some of the features depicted must be qualitatively incorrect. How Whitaker can be so confident in his ability to distinguish the qualitatively correct from the qualitatively incorrect is unclear to us; especially since there are innumerable features for which he has no plausible explanation. If his identifications of lunar features are unreliable (not due to his lack of lack of experience in recognizing them, but rather due to their being depicted in an unpredictable way), then the dates he derives based on his supposed identifications are highly questionable.

It is particularly difficult to check Whitaker's work because although he gives a photo that he says corresponds to F1 he does not tell us when it was taken (even though he begins his article by scolding Galileo for failing to do the same), and does not bother to point out what features in the photo he believes correspond to patches in the drawing. He gives no photo corresponding to F2; that is, he provides no evidence that such a total transformation of appearance can occur in two hours at this phase of the Moon. For F3 he does give a photo with certain features labeled, but again (as with all the other photos) he gives no date or time, so it is impossible to confirm that the terminator position in his photo actually corresponds to that at 16:30 UT on December 1, 1609 (the date and time he assigns to drawing F3).

The illustrations from Whitaker's 1978 article were repeated, with slight modifications but without any further explanation of their datings, in his chapter on "Selenography in the Seventeenth Century" in the 1984 Cambridge General History of Astronomy as well as in his own highly-acclaimed 1999 book Mapping And Naming The Moon. The illustrations in the latter two publications appear to be identical, and differ from the former mainly in that all Galileo's drawings and engravings are printed against a uniformly black background to make them look more like the photographs. The photographs, still unidentified, seem to be the same as those shown in the original publication. The only substantive change we can detect is Whitaker's photographic comparison with Galileo's engraving E4. In the original article, Whitaker labeled four features along the terminator, one of which is the crater Albategnius (virtually invisible in the photo). In the two subsequent publications, Whitaker retained the photo, but deleted two of the labels along the terminator ("G" for the Ukert Hills in the 1978 article seems to correspond to "D" in the later illustrations, while "J", which referred to Albategnius in 1978 has become "E") and instead labels twelve new features which he claims are clearly visible in the illuminated half of Galileo's Last Quarter Moon. Unlike the 1978 article, where all labels are explained, Whitaker tells us what only five of the twelve new labels are pointing to (and then only in his 1984 publication). We show those, all located in the Moon's heavily cratered southern highlands, in the following panel.

The arrows in the left-hand photo (which Whitaker has purposely de-focused) show what Whitaker thinks correlate with the corresponding arrows in the Galileo's engraving (center image). In the photo, Whitaker says the arrows are pointing at the craters Deslandres (d), Purbach (p), Regiomontanus (r), Walter (w), and Orontius (o). Although these identifications seem plausible at first glance, when looks more closely, any number of questions arise: why is Orontius shown so large when the photo shows it to be small? why is Deslandres depicted in the engraving when it is invisible in the photo? what is the large crater just the left of Whitaker's "E" in the engraving? As to the latter, is it Arzachel, the prominent crater just to the lower left of Whitaker's "E" on his photo? If so, why is it shown so much larger than Purbach when it is actually small? Why is it shown so close to the purported Albategnius? And what happened to Ptolemaeus and Alphonsus, the other two members of the prominent chain of which Arzachel is the southernmost member? To answer such questions one has to accept Whitaker's reasoning that Galileo's engraver has somehow shown certain craters in a "qualitatively correct" fashion (while, evidently choosing to completely omit others), but placed them in incorrect positions and grossly distorted their sizes relative to one another.

Photo Labeled by Whitaker (1984)	E4 Labeled by Whitaker (1984)	Alternative Identifications

adapted from image © Cambridge Univ. Press	adapted from image © Cambridge Univ. Press	adapted from image © Project SEE

In the right-hand image we show what seem to us an equally plausible alternative set of identifications for four of these. We have placed the arrows on a typical photo of the southern highlands, this particular one being adapted from Project SEE's Tactile Moon display, which attempts to make the most prominent features visible at various lunar phases perceptible to the blind. Their photograph was taken on September 18, 2003. The arrows now point to the centers of the following craters: Longomontanus (d), Tycho (p), Maginus (w), and Clavius (o). In this identification, the very large walled plain Deslandres (above and to the right of "p" in the photo) occupies about the right position relative to "p" to be where Galileo puts the very large crater in his engraving. Although these features seem a bit too close to the Moon's south pole in this particular photo, their exact positioning depends on the lunar librations.

We do not mean to imply that our "identifications" are any more correct than Whitaker's. In this particular alternative scheme, without invoking geometric inaccuracy we have no clear candidate for the feature to the left of the "E" in the engraving, and the object pointed to by Whitaker's "r" could be several things. Our point is that the features represented in the engraving are so vague, and real features in this region of the Moon so numerous, that any number of plausible identifications can be proposed; yet none seems to work perfectly. We suspect that if the engraver drew a completely random impression of a cratered surface we would have about the same degree of success matching his imaginary features to the real ones.

Even if we accept Whitaker's identifications as definitive, the questions do not end. Galileo's engraving clearly shows the terminator butting up against the east wall of what Whitaker says is the crater Walter, and spilling over the east walls of Purbach and Orontius. Was the terminator really at this position at 4:00 UT on December 18, 1609 (the date and time Whitaker assigns to this engraving)? The position of the terminator in the engraving is distinctly farther to the left than in Whitaker's undated photograph. Since Albategnius had already nearly faded from sight when the photograph was taken, would it have been visible at all when the terminator got to the position shown in the engraving? Would the east wall of Goldschmidt, the Ukert Hills, and the north wall of Hipparchus (three features that Whitaker claimed to have identified in his 1978 paper) have actually been on the terminator when it got to the position shown here, and looked as depicted in the engraving? What are the dots of light shown in the shaded part of the Moon, well to the right of the terminator? Could exactly this phase have even been observed from Padua in 1609-1610? We will return to these questions in a following section after we have explained how the position and appearance of the terminator for any given date and time can be predicted with great precision.

Despite the vast resources of the Lunar and Planetary Institute, Whitaker was apparently unable to find a photograph more closely matching the phase he wanted to illustrate. Our own conclusion, based on terminator predictions and comparison of them to photos taken at known times, is that there is either something drastically wrong with the way the terminator is shown in the engraving or with Whitaker's identification of these craters: by sunrise on December 18th (6:00 UT), the terminator may indeed have been approaching the east wall of Walter, but it would have been full crater-widths away Purbach and Orontius. Our alternative "identifications" are no better in this respect: it is hard to see how the terminator could spill over into Tycho before it reaches the east wall of Maginus.

In addition to the major publications described above Whitaker published a 1980 letter in the journal Science which contains illustrations comparing wash drawing F4 and engraving E3 to a photograph. Labels superimposed on F4 claim to identify three features not mentioned in either his earlier or later publications. Whitaker does not say what these three new features are, implying that the correlation between photo and drawing should be self-evident even though "size reduction and further reproduction will undoubtedly cause loss of some of the finer points of correspondence along the terminators" (modern printing technology evidently being incapable of matching that of 1610). The times he assigns to these images are the same as those given in his 1978 paper, but again he offers no proof that his undated photographs "taken at similar phases" actually match the expected appearance of the Moon at those times. Even if they did, the "correspondence along the terminators" is not at all obvious to us, neither at the points Whitaker calls attention to with his labels no anywhere else. Indeed, to us, the number of mis-correspondences seems to far outnumber the number of correspondences; although in fairness Whitaker has chosen to print the terminator in his photographs in such a way that the terminator is dark and vague (as previously indicated, to make the appearance of the visual terminator clear in a photograph the gamma has to increased to a value far larger than the one normally used). Even Whitaker admits that the regions away from the terminator do not correlate well, and that must surely lead one to question whether the terminator was really rendered with that much greater care.

Finally, in his 1980 Science letter Whitaker makes the claim that the limited accuracy of Galileo's drawings was due among many other things to his using an unmounted 20-power telescope. This assertion seems more than a little odd, since in his January 7, 1610 letter (with which Whitaker was thoroughly familiar) Galileo goes to some length to warn the unknown recipient that the telescope needs to be fixed to a stable surface to escape "the shaking of the hand."

Gingerich and Van Helden's Page-turning Theory

Gingerich and Van Helden also express the opinion that, due to the low quality of Galileo telescope, only one feature (what is now called Albategnius) showed a clear enough pattern of shading to be definitely identified as a crater (i.e., a depression in the surface), and this led to Galileo to place a highly enlarged version of it on the terminator in engravings E2 and E4. We do not feel this opinion accords either with Galileo's other observations or with his written comments about all craters showing a distinctive pattern of light and shade.

In their recent paper Gingerich and Van Helden point out that if these watercolors were studio reproductions of original sketches made at the telescope, their placement on the page makes no particular sense: they are not arranged in sequence of increasing phase, and, more importantly, some are placed with the terminator horizontal and others vertical, for no obvious reason. They proceed to tell us that this all makes perfect sense when one takes Whitaker's dates and realizes that they were recorded directly on this paper at the telescope. We, for one, fail to see how the sequence makes any more sense when viewed in this way; but what we address here is Gingerich and Van Helden's contention that the orientation of the terminator was dictated by the appearance of the Moon in the sky, with the artist choosing to hold the paper either horizontally ("landscape mode") or vertically ("portrait mode"). We admit to a little confusion since Gingerich previously told us that all the drawings are on a single stiff sheet of watercolor paper; and now he seems to be telling us that drawings F1-F6 are on a separate rectangular sheet.

We certainly agree that an artist will have a very strong inclination to record his subject in the orientation that he sees through the telescope. This contention is borne out by Galileo's recollection of the way the little stars next to Jupiter appeared on the night of January 7, 1610 as later recorded at the start of his journal page. Jean Meeus assigns a time of approximately 16:30 UT, and the angle he shows matches a modern prediction of how the moons of Jupiter would have appeared in a Galilean telescope at that time. The match would be exact if we guess Galileo was holding the paper rotated about 10 degrees counterclockwise.

The fundamental flaw in Gingerich and Van Helden's analysis is their apparent belief that the lunar terminator is always either horizontal or vertical. They say, for example, that drawings F1 and F2, made with the Moon low on the western horizon, is an accurate representation of how the lunar crescent looked when one assumes the artist was holding the paper horizontally (starting, for some unknown reason at the middle of the upper edge). In fact, the orientation of the lunar terminator in the sky is perfectly predictable at any given date and time; but it is rarely, if ever, exactly horizontal when the Moon is rising or setting.

In the following table we show accurate simulations of how the Moon would look as viewed by an observer with a Galilean telescope at the date and time assigned by Whitaker to each of the seven drawings. Below each simulation we give an updated version of Gingerich and Van Helden's presentation of how, on the assumption that the drawings were actually made at the telescope on these dates and times, the paper would have had to have been held for each night's work.

The writing on the incomplete horoscope shown in the final illustration is, incidentally, upside down. That is, whoever labeled it seems to have turned the page once again. Also, if Gingerich's initial contention that the drawings are on a single folded piece of paper with drawing F7 "across the fold" then it is a little hard to understand why, with all the blank space available, the first six drawings were crowded, out of sequence, onto one half. Note: we have since learned that the two sets of images are on separate sheets, not a single folded one, so the final illustration above (showing the sheets attached) needs to be corrected (Gingerich, 2008, private communication).

It is unclear to us what to make of this result. There is a certain consistency in the way the paper might have been held for drawings F1-F5; however, if the journal entry for Jupiter's satellites correctly indicates Galileo's propensity to hold his papers a little counterclockwise, then only drawings F6 and F7 match this pattern. Also, as previously indicated, the sequence established using Whitaker's dates still makes no sense to us. One would think that most artists would have filled the page in some orderly way: left-right and top-bottom, or clockwise (or counterclockwise) around the edge. But Whitaker's sequence follows no such pattern. The placement of drawing F4, with respect to the previous three drawings, as it would have to have been done if it was made on December 2nd, seems particularly illogical and inexplicable.

As a result, the page layout does very little to increase our confidence in the correctness of Whitaker's dates. As to whether these drawings were actually made live at the telescope or later copied in the studio, about all we can say is that if they were made live at Whitaker's dates and times the paper would have had to be held as shown. Someone more knowledgeable than us about painting might be able to tell from a careful look at the originals at the BNCF whether the brushstrokes are all consistent with an artist holding the paper as shown here. If they are, that would be fairly strong evidence in favor of the Whitaker-Gingerich-Van Helden theory of their creation. Such a person might also be able to make a judgment about the little detail of a crater that appears between drawings F1 and F2. To us this appears to be at a completely different scale (magnification) than the other images, and possibly of a different phase than any of the main drawings (it seems to show a crater partially illuminated by the Sun on a much larger radius dark limb). We have the impression that artists copying from nature usually draw things sight size, and have very little of the "zoom" capability illustrated here. The size of the image on the paper can, of course, be adjusted a little by moving the paper closer and farther from the eye, but the little cameo at a different scale makes one wonder if this was all really done live at the telescope.

Jan. 7, 1610 Letter (IMSS)	L1 - L3	C1 - C4 & L4	L5
Engravings (IMSS, Venice ed.)	E1	E2 & E4	E3 & E4
Wash Drawings (BNCF)	F1 - F6	F7
Perspective Sketches (IMSS)	M1 - M5

Folio Page 28		Folio Page 29
Back (Verso)	Front (Recto)	Back (Verso)	Front (Recto)

Adapted from original images © BNCF

F1 Nov. 30; 15:00 UT	F2 Nov. 30; 17:00 UT	F3 Dec. 1; 16:30 UT



F5 Dec. 17; 4:00 UT	F6 Dec. 18; 6:00 UT	F7 Jan. 19; 5:50 UT

GALILEO'S OBSERVATIONS OF THE MOON

Links to High-Resolution Originals

Introduction

Drawings in a Copy of Galileo's Letter of January 7, 1610

Engravings Appearing in Sidereus Nuncius

Cherrington's Inversion Theory

A Possible Printer's Proof of Sidereus Nuncius

Watercolor Drawings Bound in the Manuscript Copy of Sidereus Nuncius

Gingerich's Distortion Theory

Whitaker's Sequencing of the Drawings

Gingerich and Van Helden's Page-turning Theory