NGC/IC Project Restoration Effort

===== NGC 468 is not IC 92, as I have long supposed. Courtney Seligman wrote in March 2015 suggesting that the NGC object might be identical with NGC 472, a brighter galaxy than IC 92, and at the same declination. After looking at JH's Sweeps for a day, I wrote back to Courtney, "The short answer is that you're right: NGC 468 is almost certainly a second observation of NGC 472." Here is a somewhat condensed version of the longer answer: At first, I was as puzzled as you were by the fact that JH had apparently swept up a fainter galaxy instead of a brighter one. But this is indeed possible because of the way that both Herschels swept the sky -- they used their telescopes as meridian instruments, "sweeping" north and south in short arcs of typically 2.5 to 3 degrees, while the diurnal motion carried the sky westward across the meridian through their field of view. It's possible to miss objects if the north-south arcs are too long and the diurnal motion carries the sky too far west. I thought that this might have happened here. Wolfgang Steinicke has found cases in William Herschel's sweeps where it actually accounts for missing nebulae. With this kind of sweeping, errors in RA are just as likely as those in Dec. Indeed, in both the Herschels' observations, there are many RA errors of not just digits (e.g. 1 minute, 10 minutes, etc), but of seemingly odd amounts like 18 seconds, 46 seconds, and other strange numbers. That's the case here where, as Courtney noted, the Dec is correct for N472, but the RA is off by 37 seconds. This inspired me to dig into JH's Sweep 106 on 22 November 1827. I did a rough "reduction" of his observation w.r.t. one of the stars in the sweep, sigma Trianguli. That got me to within a few arcminutes of the position in the NGC, but not close enough to really verify JH's position. So, I had to ask how he did his reductions, a question that I've been curious about, but never had to answer before. (JH, by the way, describes his reduction technique in detail in one of his early papers on double stars. This paper, unfortunately, does not seem to be currently on-line.) After reading through many of his early sweeps -- which are preserved in a "fair copy" up to Sweep 107 so they are more legible than the actual observing logs -- I worked out what the entries in the logbooks mean. Here is the observation of NGC 468: 01 12 45 -29 2 01 10 00 The first three numbers are a clock or watch reading, the 4th is a count on a chronometer ticking off 130 (*not* 120) "beats" per minute, the fifth is the eyepiece wire -- there were two of them (though JH also occasionally used the trailing edge of the eyepiece field as a third "wire") with which he timed the meridian passage, and the last three numbers are the NPD "index" on a scale marked in degrees, minutes, and seconds of arc. To reduce the RA, JH applies various corrections to these measurements, the main ones being the clock correction for each wire, along with the difference between the transit and the count of the chronometer "beat." (All this, incidentally, must have been recorded by JH's assistant, usually John Stone, finally mentioned by name in the Cape of Good Hope Observations, so that JH would not have had to ruin his night vision.) The clock corrections, as well as the index zero points are worked out by using known star positions, of course. For this sweep, JH has clock corrections of -1m 37.0s for wire 1, -2m 13.5s for wire 2, and -2m 45.9s for the trailing edge of the field. Applying the number for wire 2, along with 13.4 seconds for the 29 chronometer "beats" to the 01 12 45 clock reading gives an RA of 01 10 18.1 for the date of observation (in this case, JH simply took 1828, close enough given his usual accuracy). Similarly, the mean NPD zero point is +57d 02' 36" for the sweep, so the NPD is 58 12 36, again for epoch 1828. JH then precessed this to 1830 with his precession tables (which I haven't uncovered yet), but using the modern precession, these come close to the numbers that JH included on his reduction pages, that he copied into his published papers, and that Dreyer eventually copied into the NGC. Going through the reduction again shows that there is no large error in JH's calculations or the resulting positions, in either RA or NPD. But, looking at the log, I noticed that the "2" notation for the wire number appears to have a "1" written over it. It's not very clear, and could well be the result of the photocopying process as the RAS people were copying the page for the Herschel Archive. But the "-29" has a similarly overstruck "2", so I wondered what would happen if the wire number actually was "1" and not "2". Going through that reduction gave me the RA as 01 10 54.6, with the same NPD, 58 12 36, for 1828. This becomes 01 20 30.2, +32 41 45 for J2000. Compare this with the modern value of 01 20 28.7, +32 42 33 from UCAC for NGC 472 -- there it is. So, it looks to me like the wrong wire was recorded in the Sweep. If this is the only mistake, it seems reasonable to simply adopt JH's published position and correct it by the +36.5 seconds of time between the first and second wires at the same declination. The agreement in position between the two observations is even closer with this easy fix. All this makes additional sense if JH actually saw the brighter NGC 472 rather than the fainter IC 92. We can't be absolutely sure that this is the correct explanation, of course, but this looks like another kind of "digit error" that could happen fairly commonly. I will re-examine other of JH's apparent RA errors to see if the same reasoning can recover other "lost" or misidentified objects that he observed. (A couple of additional cases of possibly mistaken wires are the clusters NGC 6396 and NGC 6400, both of which see. Their identifications are not in doubt, but the RA's are off.)

===== NGC 468 See observing notes for NGC 472.