Observing List of
"Not Found" Objects

Good hunting! - Bob Erdmann

The galaxy may not be irretrieveably lost, however. Since the declinations in the first two Leander McCormick lists are generally (though not always!) reliable to within a couple of arcminutes, it may be possible to scan around the sky at Leavenworth's declination to find the object (see e.g. NGC 331). I haven't tried yet, however.

In fact, I can't find Leavenworth's object anywhere near his position. Nor are there any other nebula/star pairs within several degrees of that position that match the sketch, either. The galaxy chosen by ESO, 3.8 minutes preceding and 19 arcmin south of Leavenworth's position does not match the sketch, so that cannot be the object, either.

Leavenworth added a note "Neb?" to his description, so it is possible that the object is simply a star. However, I could not even find two stars in the correct relative orientation in the area that would match the sketch.

The sketch is dated 15 Oct 1885. Leavenworth made at least four other sketches that same night. They are of N377, N540, N635, and N842 (all of which see). Of these, N540's identification is unsure, and N635 is three degrees south of its nominal position. Assuming all four identities, though, the average offset of Leavenworth's positions in RA is +25.3 seconds of time with a mean error of +-32.2 seconds, and a standard deviation in one observation of +-64.5 seconds (all are at roughly the same declination, so the conversion to arcseconds can be ignored given the size of these numbers). In Dec, the equivalent numbers are -5.3 arcmin, +-4.2 arcmin, and +-8.4 arcmin. Given offsets and errors of this size, and the three-degree accidental error for N635, NGC 412 could be ANYwhere within several degrees of Leavenworth's nominal position.

But I still can't find it. So, unless other folks want to spend more time on the field, NGC 412 is probably irretrieveably lost.

Dreyer has a curious statement in his note in the Scientific Papers (1912). Citing the observers above as having "... seen only one nebula," he goes on with "no doubt the following one." Yet all the observers have assigned the preceding number (H III 154 = N420) to the object. Dreyer himself followed JH's lead in this, giving the earlier number to the object that JH, d'A, and LdR all saw.

In any case, there is no nebula in the area that might be N421. Since assigning the number to one of the stellar objects mentioned above is pure speculation, I'm not going to do it. Thus, N421 is "Not found."

Wolfgang Steinicke again drew my attention to this missing pair in July 1998. I made a further search at "reasonable" digit errors (e.g. 1hr in RA, 10deg in Dec), but found nothing matching Muller's description anywhere near any of the resulting positions. It may be worthwhile for other interested investigators to cover the areas, too -- they may have more luck than I.

There is a faint knot (or superposed companion) in the corona of N932, but it is only about 35 arcsec northeast of the nucleus. While Copeland might have been able to see this, there is no way to make his measurement fit. There are no other likely galaxies nearby that he might have seen, either -- aside from NGC 938 about 10 arcmin east-southeast which he, in fact, also saw.

So, NGC 930 is a mystery.

The modern catalogues, by the way, are wrong in adopting that number for the galaxy that is here. Dreyer clearly meant NGC 932 to apply to WH's object.

The next thing to try is to look for galaxies in the area that are at the offset inferred from Stephan's published positions. These are -4m 25.61s in RA, and +2m 58.0s in Dec. A cursory scan of the relevant areas didn't turn up any reasonable candidates, but I suspect that a careful inspection of the fields northwest of the stars between 5 and 9 in Triangulum would eventually reveal Stephan's object.

Until then, however, N952 is unfortunately "Not found."

After I wrote the preceding paragraph, I learned that Wolfgang Steinicke (and others) have suggested that NGC 1037 is actually UGC 2119, two minutes of time preceding Swift's position, and 6.7 arcmin south. This is certainly possible as there are several other larger RA errors in Swift's 5th list. However, this still leaves the problem of NGC 1032 being nearly 3 degrees to the north.

Looking at the field, two other possibilities suggest themselves. First, Swift may have picked up UGC 2106 which is in the same field as UGC 2119. This would suggest that he somehow thought that U2119 was NGC 1032.

Secondly, if he had NGC 1032 correctly identified, then it is just barely possible that he might have also seen the very faint galaxy about 4 arcmin northwest. This is quite flattened, and might be visible in a 16-inch under very good skies. However, there are brighter stars near it -- in particular, a star is less than an arcminute to the northeast. Why didn't Swift mention any of these? This hypothesis also requires a large error in position (50 seconds in RA and 2 deg 53 min in Dec).

All in all, I'm not convinced by any of these hypotheses, so shall stick with "Not found."

I haven't yet done the library work to know if the comet's tail stretched off to the northeast from the head. But the description makes it possible that this is the correct explanation for this NGC entry.

Whatever the case, there is nothing in Bigourdan's position, which comes from two accordant measurements on 17 December. About 40 arcsec to the southwest is a faint double star that he probably could not have seen (based on the fact that he had difficulty with NGC 1177). He adds a curious note to his description: "At the end of the measurements, I could see the object very well: the sky, very clear at just that moment, had been a little unsettled." This is what leads me to believe that he has misidentified his star field.

A search of the surrounding POSS1 fields turns up no digit errors in the ten's places of RA and Dec that would nail an appropriate object. So, this object may well be lost. Wolfgang's identification, by the way, is a star about an arcminute west of Swift's position. Swift's description pretty well rules out this ID.

However, the center of the Fornax cluster is one and a half degrees north. I think it's likely that Swift saw two of the galaxies there, but choosing two out of the 15-20 that he could have seen would be pure guesswork. Similarly, Lauberts's guess in ESO (ESO 358-G040) is based on a reliance on the 1 degree difference in declination more than it is on the likelihood that Swift actually saw the object: ESO 358-G040 has a total visual magnitude around 16.2, likely putting it beyond Swift's limit, especially given the far southern declination.

A third possibility is raised by Kreutz in a note following Swift's list in AN. Kreutz notes that the offset of Swift's position from that predicted for the comet by Finlay is 38 minutes east, and 4 degrees 1 arcminute south. However, searching at Finlay's place for a double nebula turned up nothing, either.

Other objects found by Swift on the same night include NGC 1797 and NGC 1799, both very near Swift's positions for them; and NGC 2589, like NGC 1392, not found at Swift's position (see Herbert Howe's note in MN 61, 29, 1900, copied into the IC2 Notes).

In the end, NGC 1392 is another of Swift's nebulae "not found."

In any event, N1408 is currently unidentified. See NGC 1378 and NGC 1396 for more on Schmidt's Fornax Cluster work.

The description of Leavenworth's "2nd of 2" matches the one galaxy in the area. But that is it. The star 15 sec west of the galaxy is 13th magnitude, and there is an 11th magnitude star half that distance northwest. Why did Leavenworth not mention that? I think that Leavenworth has misidentified the known galaxy so that his description applies to a different pair altogether.

But which pair? I don't see any other in the area that matches the descriptions. So, NGC 1429 is another lost NGC object until someone with sharper eyes than me has a go at the problem.

Among the more likely candidates are NGC 1599, and RNGC 1610. N1599 is at the same declination and is just over a minute of time west of the nominal place of N1610. However, it has a bright star just 1.1 arcmin east-northeast which Leavenworth would almost certainly have mentioned had he seen this galaxy. The RNGC identification is also possible, but is 7 arcmin south and 1.4 minutes of time east of Leavenworth's place.

Neither of these options is particularly compelling, but are still worth noting.

Again, as with N1610, there are two candidates for N1619 that seem more likely to me. RNGC 1610 is two minutes of time west and 8 arcminutes north, while the 51st object in Reinmuth's 1932 list of nebulae is 1 minute 20 seconds west and 28 arcmin north. I'm not convinced that either of these is Swift's lost nebula, but I note them in any event.

Delisle Stewart searched for N1655 on a Harvard plate, and could not find it either. He has an intriguing note attached, however: "... a hazy star p 1 minute, same Dec." I don't even see that "hazy star" on the POSS1 prints; is it a defect on the Harvard plate?

So, N1655, too, is presumeably lost.

As with NGC 1655 (which see), which Lohse found in the same part of the sky, there is no trace of these at Lohse's given position. Three arcmin south is a group of six faint stars about an arcmin across, but we would need visual observations to tell if these could be mistaken for two nebulae.

Lord Rosse and his observers tried six to eight times. The actual number is uncertain. In the 1861 monograph, LdR says "Looked for seven times. Not found," while the 1880 monograph says "Looked for 6 times in the years 1848-58, not found", and adds another attempt on 30 December 1867: "Night bad; suspected slightly luminous appearance in spots, but not very decided." (Dreyer in the NGC note says six times). Neither Tempel nor Spitaler could find it, and Reinmuth did not locate it on the Heidelberg plates.

I searched for this on the sky survey prints and later on the DSS, as well as in the IRAS infrared images. The closest nebulosity is over 20 arcmin from JH's position and does not agree at all with his description. In any case, the whispy stuff is far too faint to be seen visually.

Even JH came to doubt his observation. His GC note reads, "A very large diffused nebulosity, distributed in zigzags. This has been looked for seven times by Lord Rosse and not found. Its existence is therefore very doubtful."

WH places the nebulosity 1 min, 26 sec east, and 7 arcmin south of Eta Ori. There is nothing here on either of the POSS1 plates, nor on the SERC EJ plate. However, 7 arcmin north of Eta Ori there is a very faint, very diffuse sheen of nebulosity (I make the approximate position 05 23.0, -02 20). But could WH have seen this? I very much doubt it. So, I've tentatively labeled this "Not found."

Also, JH originally made this an observation of his father's V 38: that identity is shown in his 1833 catalogue. However, his position differed enough from WH's that he made them separate objects for GC; Dreyer followed suit for NGC. There, H V 38 = NGC 1909 (which see).

In short, NGC 1927 is "Not found."

I see nothing in the field on the POSS1 prints that looks like a cluster. Perhaps a visual observation can turn up something.

Herschel's description might just as well fit UGC 3511 (06 38 45.8 +65 15 22, 1950), a rather patchy late-type spiral galaxy, but the position is off by random amounts in both coordinates. Similarly, the CGCG object at 06 38.2 +65 43 (1950) is probably not WH's object.

Since there are no reasonable solutions that we can easily see, we'll just have to let NGC 2253 be "Not found" for the time being.

At the moment, my feeling is that Tempel confused NGC 2633 with NGC 2629, and that his pair is actually NGC 2634 and NGC 2634A. These two galaxies are the only ones in the group that are close enough to be actually called a "pair." However, while N2634 is bright enough to rival the earlier observers' discoveries in the area, N2634A is certainly not. It's just conceiveable, however, that on a night of exceptional transparency, the pair may have stood out enough to capture Tempel's attention. He was, in fact, so struck by their brightness -- compared to the nearby nebulae that the Herschel's and d'Arrest found -- that he suggested variablility for them.

This is a pretty weak argument, however, so until Tempel's discovery sketch (which he mentions explicitly) can be examined, the question of the identities of these two NGC numbers has to remain open. So, I've simply entered the NGC positions in the table for the time being.

"Date       GC     RA  (1860.0)  Dec      Remarks
1870 Mar.24  --  09 54 19.6  +60 47 58.2   G.C. 1998 s f neb; p 45 deg
                                            s 2'. [Place only approximate.]
1870 Mar.24 1998 09 54 36.7  +60 46 33.3   G.C. 1998:  F; S; R; mbMN."

There are three things to note about these observations:

  1) The position of the second (GC 1998 = NGC 3102) is from the GC.
  2) The "p 45 deg s 2' " means that the first (N3097) is 2' away
     from GC1998 at a position angle of 45 degrees.  This is 
     inconsistent with the position which implies the object to
     be northwest, not northeast, of N3102.
  3) Both observations are credited to E. P. Austin, and there is 
     a note for N3097:

     "Perhaps a nebulous star.  It is half-way between G.C. 1998 
       and a star 11 magn."

The positions don't tell us anything we don't already know since they are correctly transfered into NGC from GC and the Harvard list.

Since Austin was observing with a 15-inch telescope, I don't think that he could have seen either of the faint stars Glen Deen measured during his MicroSky project. The magnitude estimate given by Austin for the "star 11 magn" is rough since there is nothing that bright near the galaxy.

WH had this to say (N3102 = H III 916): "eF, vS, Stellar. Near a S st." And JH: "F, vS, R, bM; a coarse D * nf points to it; has a * 11 30'' dist, pos 142.2 deg ." All of JH's stars are identifiable, and I think that his star 11 must be the same one mentioned by WH and by Austin.

So, where does that leave N3097? My guess is that Austin has misidentified another nebula as N3102, but I don't see it or its purported companion in the area. A more thorough search may turn them up.

Between five to ten arcmin southeast of the position reduced from the N3423 offset (10 46 45, +06 09.5; B1950.0), there are one or two asterisms of stars that WH might have picked up. The positions are far enough off, however, that I doubt these stars are WH's object.

Dreyer searched for this on the Greenwich plates that he asked to have taken of the area covered by one of WH's very strange sweeps (see NGC 2938 and NGC 3752 for more). I've searched for it on the POSS1 prints. There are no candidate galaxies within 30-40 arcmin of JH's position. So, we just have to take JH's word for it -- "A very doubtful object," indeed!

I searched at all the reasonable digit error offsets where I've found other "lost" nebulae, but found nothing that resembles the north-south pair in the table, nor the galaxy flanked by stars in the sketch.

So, unfortunately, two lost nebulae.

Since there is still nothing at Bigourdan's place (assuming the mistaken identity for the comparison star), I'm tempted to assume some kind of error in his observation leading to the digit errors. But so far, I've not been able to find it.

There is only one galaxy here, N4310, and that was also seen by WH, as well as by JH during the sweep previous to the one during which he saw two objects in this place. It is possible that JH misidentified another pair -- but there is no other pair near his place, nor at any reasonable digit error from his place. In addition, his measured position for N4311 is less than 20 arcsec from the single galaxy here. This is would be a remarkable coincidence if the position actually applies to another object.

So, we are confronted with another lost NGC object.

A possibility for WH's object is NGC 4310 = NGC 4338 (which see). It is at roughly the correct RA (17 seconds preceding WH's), but is 1 deg 50 arcmin south of WH's Dec. This makes it unlikely that this is the object he saw.

Finally, Reinmuth, RNGC, and Steinicke have called N4317 a star. I think this is unlikely as WH would have probably noted the object "very small" or "extremely small" rather than simply "small."

Checking the signs of Tempel's offsets turned up nothing that matched in any of the other three possible positions. There is a double star (noted in the position table) that might be his object. It has a faint double galaxy about an arcmin to the northwest, and an even fainter double star a bit further away to the southeast -- these may have enhanced a nebulous appearance a bit. However, adopting the brighter double as Tempel's object would require not only changing the sign of the declination offset, but its size and the size of the RA offset as well. So, I doubt very much that the double is Tempel's intended object.

Until Tempel's original observing records can be examined for possible errors, this object will have to remain "lost."

The other three were seen only once by Lord Rosse, in 1865. He has a sketch showing their relationship with the surrounding star field, so it ought to be easy to identify them. His sketch, however, bears no relationship to the sky around NGC 4914, so it seems likely that he misidentified the main galaxy. Dreyer took this possibility into account by questioning the identification of the main galaxy in the descriptions of N4912, 4913, and 4916.

A search of the four POSS1 fields around N4914 also turned up nothing that matches the sketch. Another possibility is that LdR made a 10 deg error in his position, and actually observed several galaxies in the Coma Cluster. But that is about 10 deg 25+- arcmin south, and there are no galaxies/stars in the cluster area that match his sketch. One hour errors are also possible, but I've not yet looked closely at those.

The next thing to do will be to look into the lists of nebulae that Lord Rosse had at his disposal at the time (e.g. JH's 1833 observations, the GC, d'Arrest, Auwers) to see if any objects listed near NGC 4914 might be the one that he observed. Since the pattern in his sketch is clear (3 of the objects in a north-south line with a 4th following the southern most), it should jump out at us when we see it. I hope.

Wolf claimed to have found and measured NGC 4912 and NGC 4916 on his plate with the IC objects from his fifth list. The object he took as N4912, however, is a star, and his N4916 is a defect (which he did not mark) on the plate. There is nothing in its position on POSS1.

Since there are no galaxies in the area matching JH's description (all are too small), nor are there any one hour preceding or following, or within two+- degrees of the nominal declination, this may well be a visual illusion of some sort, perhaps caused by scattered light in his telescope.

However, Swift saw a star "very close" to the object. None of the galaxies has stars nearby that could be described that way, and the positions are well off Swift's nominal position. So, this object is probably lost.

I searched at reasonable digit errors (+- 1 deg and +- 1 minute) with no luck, so the best we can do with this object for the time being is "Not found."

IC 4533 is also unlikely to be the object Swift saw because there is a brighter star just a couple of arcminutes northeast of the galaxy; Swift would probably have mentioned the star in his description, as Javelle in fact did.

The IC1 note is another curiosity about the field: there is nothing at Bigourdan's "revised" position, either. There are a few faint stars and even fainter galaxies in the area, but nothing that Bigourdan could have seen. He comments under his observation of IC 4589 (which see), that he doesn't see how it would be possible to mistake IC 4589 for N6059. He's right.

Finally, there are no systematic offsets in Swift's positions for the night of 6 May 1886 that might lead us to a galaxy. Another (N4280, which see) of those objects, though, is also probably lost.

Whatever the case, these are two of a group of four nebulae that Lewis Swift found on the night of 28 June 1884; the other two are NGC 6232 and NGC 6236. Over a year later, on 11 August 1885, Swift found another nebula, NGC 6248, about half a degree south of his group. There were no other observations before Dreyer compiled the NGC, so he included all five.

Looking at the area on the Sky Survey prints, we now see only three galaxies here that are bright enough that Swift could have seen them. These are NGC 6232, 6236, and 6248. Swift's RA's for the three are systematically too small by 20 to 25 seconds of time, but his declinations are very good. Looking at his positions for the missing two objects shows that the declination of NGC 6237 is close to that of NGC 6232, and that for NGC 6245 is similarly close to that for NGC 6236. In addition, his RA's for the two missing objects each have roughly the same offset from the RA's for the same two galaxies (32 seconds in the first case, 48 seconds in the other).

So, I wonder if NGC 6237 = NGC 6232 and NGC 6245 = NGC 6236 -- in spite of the fact that Swift found all of the objects on the same night, and explicitly noted "1st of 4," "2nd of 4," etc, in the descriptions of all four objects. Keep in mind his method of finding positions: centering the object in the eyepiece, and reading the setting circles. Did he perhaps bump the telescope or setting circles inadvertantly after reading positions for the first two objects?

Still, he used a very large field eyepiece, so it may be that he mistook stars near the galaxies as other nebulae. Or, he may have seen reflections of stars out of the field and mistook them as nebulae. Or, his eyes may have played tricks on him if he was tired. I favor the jarred telescope/setting circle hypothesis, but would not bet even a nickel on its being right.

Whatever happened, the two objects do not exist, so I've simply entered them as "Not found" in the table.

There are several galaxies in the area that Swift could have seen, but none matching the pattern described by him. Howe also could not find the object, though he actually searched for it three nights, not just two as in Dreyer's IC2 note.

Unlike most of Dunlop's nebulae, JH claims to have seen this one, though only once. He lists an estimated position that is close to Dunlop's and calls it "A large milky way patch, much compressed, one portion much more so."

However, checking the position on the SERC IIIa-J film shows nothing more than a rather unremarkable part of the Milky Way. Nothing stands out that strikes me as something that would catch an observer's eye. Compare this to other Milky Way fields that have NGC numbers (e.g. NGC 6476 and NGC 6480) from JH's sweeps -- there is nothing obvious here. I've put the nominal position in the table.

I also checked the other nebulae seen in the same sweep; all are at about the same declination, so there is no large error in that part of JH's observation. A large RA error is possible, but I found nothing in the obvious places (plus or minus one minute, ten minutes, etc.).

Perhaps a visual observer can turn up something here.

Are there any other candidates in the area? Galaxies that could be force fit to Swift's description include NGC 6505 (= UGC 11026), NGC 6536 (= UGC 11077), and CGCG 322-032. None of these, however, are at positions that would be even digits off of the nominal position. I don't think they are likely to be the correct nebula, either.

I'm listing the CGCG identity with a question mark. It's clear to me that it is the wrong object, but there is nothing else that comes as close.

Bigourdan's single observation a decade after Edward Swift's is for an asterism of five stars, the brightest three in a line extending from northwest to southeast. I don't think that this is likely to be Swift's object, but it is a possibility. The asterism is 20 seconds east and 2.5 arcmin south of Swift's position, but I don't think that it would match his description. This could be easily checked, of course, with a 15-inch class telescope.

Of the nine other objects that Marth found the same night ("1864.59"), two bracket N6693 in RA, and are at similar Dec's: N6688 and N6713. Neither has a large offset in Marth's position from the modern positions, so I have to presume that N6693 is also unaffected by any systematic error.

Barring a large digit error (e.g. 1 degree, 10 minutes), Marth's object is probably gone forever.

Barring a digit error, though, this object may be gone forever.

The implied offsets (for equinox 1870.0) from the star are -4m 18.53s and -9 arcmin 25.5 arcsec. Not only is there nothing at these offsets from his nominal star, I find nothing at similar offsets from other stars in the same area of sky.

Unfortunately, the object is not listed in Esmiol's 1916 collection and re-reduction of Stephan's nebulae, so unless Stephan's original observing and reduction logs can be found, we will probably never recover this object.

But no nebulosity or asterism exists at Stephan's position, or at the positions implied by sign errors in the offset. Furthermore, a search of the POSS1 prints shows no nearby star with an obvious nebula at the correct offset. Jim Caplan finds no trace of NGC 7054 in Esmiol's 1916 monograph, so this object has to be listed simply as "not found."

About half a degree on to the northwest of the NGC position is a 1.5 degree long streak of interstellar "cirrus", dust well above the plane of the Galaxy reflecting the light of the Galactic disk back to us (see IC 336 for some of these dust clouds that are definitely in the catalogues). The cirrus is most easily visible in the IRAS 100-micron ISSA images of the area, though it also shows up in the 12-, 25-, and 60-micron images, and on the POSS1 prints, AND on the IIIa-J film copies of the latest optical surveys from Palomar and Siding Spring.

While I suppose it is just vaguely possible that this may be Baxendell's object, his clear description of a southern boundary just 7 arcmin north of M2, and of a nearly round shape, almost certainly rules this out. The IRAS 100-micron images in particular show a "hole" in the dust north of M2, just the opposite of what we'd expect if the nebula were real.

My own feeling about this object is that it may have been a reflection of some other object (perhaps even M2) within Baxendell's telescope or eyepiece, and that later observations are similar illusions simply "wished" into existence (see NGC 2529 and NGC 2531 for a discussion of two other such objects).

Also see NGC 1990 where an apparently similarly illusory nebula has been seen around a bright star.

For the record, that description reads, "eF, R, bM, ill-defined; a vF double star 45 deg np 4 arcmin dist points just to it." I scanned the POSS1 prints for several degrees around JH's nominal position, but found no galaxy in the area with a faint double where JH placed it.

So, even with two positions and a striking description, the object remains at large.

So, lost.

The nearest object fitting his description that he could have picked up is NPM1G +12.0573, chosen by RNGC and Wolfgang Steinicke to carry the NGC number. However, it is 40 seconds of time preceding and 7 arcmin north of Marth's position, not an obvious error to make. Another candidate is CGCG 430-021 -- but that is even further away at 2 minutes 45 seconds of time preceding and 5 arcmin north.

My own desperate, last-ditch, guess is that Marth picked up one of the faint stars nearer his position, but I have no idea which one. He found ten other nebulae that same night, but there is no significant systematic offset in his positions for them from the modern positions, and all are within 1.5 arcmin of his nominal positions. So, N7405 stands alone among them as unrecoverable.

I have not seen the entry in the Markree Catalogue, so do not know if it or the NGC entry might contain a typo. In any event, there is nothing at all near the place, not even a star bright enough to be included in the Markree list.

Wolfgang chose a 19th magnitude galaxy 3+ arcmin southeast of the nominal position. That, however, is too faint, has no 10th magnitude stars 20 seconds preceding, and has a different position angle; it cannot be the object that Muller saw.

I have not checked large digit errors (10 degrees, 1 hour, etc). Someone with more time and patience than I might uncover Muller's nebula that way.

A search of the POSS1 prints around the nominal position reveals no galaxy matching Stone's description. Since there is no sketch, and Stone mentions no nearby stars, we probably won't be able to recover this object.

As with NGC 7471, I have not checked for large digit errors.

There is a fairly rich, scattered group about a degree north of JH's "very loose" position. Just about any of the brighter of the galaxies in it could be the one he saw, with the some of others being his "suspected" nebulae.

Just to be sure, I checked for other objects found in the same sweep; there are only two, N14 and N7810. JH's positions for both are well within an arcmin of the modern positions, so there is no reason to suspect a systematic offset in the position of N7555. There certainly is, however, an accidental error.

See NGC 7667 for more about these observations.

In 1866, he published in AN 1571 a short list of fourteen new nebulae that one of his fellow Jesuits, Brother Ferrari, discovered during a (fruitless) search for Biela's Comet from 11 November 1865 to 18 January 1866. Father Secchi has this to say about the 9.5-inch Merz equatorial at the College Romain, "From this study, we have convinced ourselves that the refractor at our observatory is at least as keen and powerful as the Herschels' telescopes ..." (translated by me from his French original). Also, he says that they "fitted [to the telescope] a large eyepiece which gives a 27 arcmin field." (My thanks to Wolfgang Steinicke for digging out the size of the telescope.)

Back to the fourteen new nebulae. I searched near the nominal positions on the POSS1 prints for all of these, and was unable to find any trace of eight of them (NGC 7565, N7613, N7614, N7666, N7667, N7668, N7669, and N7670). There are good candidates for three others (N7683, the only one of the objects whose position was determined by actual comparison to a star, N7738, and N50), and poor candidates for the remaining three (N7663, N7739, and N116). Secchi (or Ferrari) also "corrects" WH's positions for two nebulae, N157 and N7648. His positions for those are indeed better than Herschel's -- but they don't help us find the other missing objects in his list.

If we take mean offsets from modern positions for the "good" candidates -- excepting N7683 for which the position comes from a different method -- and the two corrected WH galaxies, we find systematic offsets of -5 seconds of time, and -1 arcmin 10 arcsec in Dec. The standard deviations on these numbers (+- 18 seconds and +- 26 arcsec) suggest that the RA offset is not significant and that the Declination offset is barely significant. But even that does not help us find the missing objects.

Reading more of Father Secchi's note, I learned why the positions are so bad. "The position is determined from the setting circles of the equatorial, corrected for instrumental errors, simply by placing the nebula in the center of the field." Secchi, however, also says that he verified each of the nebulae after Brother Ferrari found them. He must have done this on the same nights as their discovery since he never would have recovered them otherwise. Since Secchi gives no equinox in his note, I, like Dreyer before me, have assumed that his positions refer to the equinox at the date of observation, i.e. 1866.0 give or take a few weeks. I adopted 1866.0.

Specifically for NGC 7667 and its cohorts: there is nothing at all near the single nominal position that Secchi gives for them, and only one or two of the galaxies within a degree of that position are bright enough to have been seen with a 9.5-inch telescope.

However, Steve Gottlieb has suggested that some of the knots in the arms of UGC 12578 might be N7668, N7669, and N7670 which Secchi says "surround" N7667. These are much too faint for a 9.5-inch telescope, but the galaxy itself is quite bright enough to be one of Secchi's objects, in spite of having a pretty low surface brightness. However, it is 3 minutes off in RA and nearly five arcmin in Dec from the nominal position, so it would be a stretch to point to this object.

There are also three other objects within 13.5 arcmin of it that might be Secchi and Ferrari's other three nebulae: UGC 12589, and the double stars at 23 21 54.6, -00 12 35 and 23 22 12.5, -00 21 42 (1950 positions). All are northeast of U12578, though, and Secchi's description clearly translates as "Very faint: the other three surround the 9th [in the list = N7667] in the field." So, U12589 and the double stars are pure guesswork, and I don't think that I'd want to stake my life on them -- or even on U12578 being N7667.