NGC 6902

NGC 6902 may also be IC 4948.  See IC 4946 for the story. - Dr. Harold G. Corwin, Jr.
====================================================================================
IC 4946 is probably identical with Shapley-Ames's "New 5", ESO 285-G007.  
Swift's declination and description fit pretty well:  "eF, S, R, 3 or 4 sts f,
form with the neb, a circle; sp of 2"; the stars are there.  (The "nf of 2" is 
IC 4948, which I'll write about in the next couple of paragraphs.)  Swift's 
position is 20 02 30, -44 10.9 (for 1950); there are no nebulae near that 
position that Swift could have seen.  He dates his discovery to 11 Sept 1897.  

The position of the galaxy is 20 20 31.6, -44 09 28.  This leads to the major  
problem:  Swift's RA is a full 18 minutes of time off.  However, the presence 
of the other object which Swift noted gives us a chance to test the 
hypothesis.  There is indeed another galaxy in the relative position given by 
Swift's observations, NGC 6902.  If I4946 is indeed New 5, then I4948 is 
identical to N6902.

Here is what Swift has to say about I4948:  "vF, vS, R, bet a wide D * f and a 
* np; nf of 2".  His position is 20 02 59, -43 50.9 (1950); that for N6902 is 
20 21 02.2, -43 48 57.  Again, the RA is 18 minutes out.  Swift gives the 
discovery date as 17 Sept 1897, a week later than for I4946.  

How could he make the same unlikely 18 minute error on two different nights?  
I wonder if it is possible that he got his date wrong for I4948.  He has no 
other objects recorded on 17 Sept, but there are two others on 11 Sept 1897, 
I4998 (which see), and I5018 (but these may be identical -- more confusion!).  
If "17" is a transcription error for "11", then Swift's observations make more 
sense.  It's possible, however, that, on the 17th, he zeroed his setting 
circles on the galaxy that he found on the 11th.  In this case, his relative 
position would be good (as it is), but his absolute position would be off once 
again by the same amount.

Also, Swift's description of the star field around I4948 is not a good match 
to the sky.  It is, in fact, a better match to the stars around I4946, 
particularly the "wide D * f" (these two are the brightest of the circle of
stars he notes for I4946).  This leads me to suggest that I4948 may possibly 
be a duplicate observation of I4946, this time with a large Dec error as well 
as an even larger RA error.  This isn't very likely at all, of course, but 
given the problems here, even this may be possible.

Finally, adding to the mess is one of Delisle Stewart's nebulae.  He did not
number it, but gave the discovery credit to Swift.  Dreyer followed Stewart's
lead, and included both observers in IC2 for I4948.  Stewart's RA is the same 
as Swift's, but his declination is 5.3 arcmin north (Dreyer adopted Stewart's 
Dec).  His description reads "F, S, R, bM" from a one-hour Bruce plate (number
3701).  There are only stars in Stewart's position -- but, interestingly, a 
wide double star follows it by an arcminute or so.

If this were the only observation of the nebula, I'd say that Stewart has 
another plate defect (see I4922, I4924, and I4940, the only other objects 
that Stewart found on plate 3701; all are probably defects).  Also, he does 
not mention the double star, and I at first took it to be his object.  In 
fact, the double could well be his object, but we will need to examine the 
plate to know for sure.  At the moment, though, it looks like Stewart's I4948 
is indeed another defect or perhaps the double star.

So, that's the evidence.  I'm leaning toward the 18 minute of time error for
both objects, but there are enough pieces of contradictory evidence that I 
can't be sure about either one.  

In the end, I've marked the I4946 identification with colons, and that for
I4948 with question marks.  That about sums it up.
 - Dr. Harold G. Corwin, Jr.
=================================================================
IC 4946 and IC 4948.
Swift List XII, #15.  20hr 02' 29.905s - 44 10' 54.048" (1950).
20hr 05m 59.185s - 44 02' 16.676" (2000). (IC 4946).  
Swift List XII, #16.  20hr 02m 59.184s - 43 50' 52.182" (1950).
20hr 06m 27.750s - 43 42' 13.008" (2000). (IC 4948).
This is an extremely interesting problem involving a large RA error which has 
evidently misled a number of the modern authorities.
Swift describes his #15 as "eF, S, R, 3 or 4 st f, form with the neb, a circle, sp 
of 2." and his #16 as "vF, pS, R, bet a wide D* f and a * np, nf of 2."  It is 
important to know that from his positional data the separation values between 
these two objects would be 29 tsec RA and 20 arcmin declination.
When his coordinates are entered into the DSS neither one comes up with a suitable 
nonstellar image.
Stewart in his Harvard List identifies as Swift XII, #16 a nebula which he places 
at 20hr 03.0m - 43 25'.7 and it his coordinates that Dreyer employed for the 
identity IC 4948, however, again when these coordinates are applied to the DSS no 
suitable image is found.
The modern authorities have the following identities:
The NGC 2000 and MOL each place IC 4946 at coordinates based upon Swift's given 
position.  The APL, RC3, Steinicke PGC, SIMBAD and NED identify the galaxy ESO 285-
G007 at 20hr 20m 31s - 44 09' 30" as being IC 4946, although the ESO itself does not 
make this equivalency.
The NGC 2000 and MOL each place IC 4948 at coordinates based upon the Stewart 
correction.  The NED also selects a 17.32 Mp galaxy about 50 tsec following the RA 
as given by Stewart and identifies this as being IC 4948.  Meanwhile both the APL 
and Steinicke identify the galaxy ESO 285-G008 at 20hr 21m 02s - 43 48'.9 as IC 
4948 = NGC 6902, which the PGC, SIMBAD and the ESO itself identifies only as NGC 
6902 without any equivalency involved.  The RC 3 also gives only the identity NGC 
6902.
Thus what the APL, Steinicke and NED suggest is IC 4946 would require a RA error 
on the part of Swift of 18 tmin, an unusually large error even by Swift's 
standards and at first I was reluctant to accept their suggested candidate, 
however, further investigation produced some findings that strongly suggest that 
Corwin's identities in the APL are indeed the correct ones.
When the image of ESO 285-G007 is examined on the DSS the stars referred to in 
Swift's description are found as described and this is also true for NGC 6902 = 
ESO 185-G069.  Also the brightness of both of these ESO identities is such that 
they would have been within the capabilities of Swift's telescope even at such a 
low declination, whereas the NED candidate for IC 4948 at 17.32 Mp is obviously 
far too faint to have been seen by Swift.  However, the evidence that for me 
confirms Corwin's identities is that the separations between the two candidates he 
gives are 29 tsec RA and 20.6 arcmin declination, which can be considered an exact 
match with the separations as given by Swift.
It would appear that Swift's published RA for both identities result from his 
having misread his instruments as the 18 tmin error does not strike me as being 
due to a typographical error, and that IC 4946 is ESO 285-G007 and IC 4948 is 
equal to NGC 6902 (h3827) and the full credit for correctly identifying both 
galaxies belongs solely to Dr. Corwin.
NOTE:  Since writing the above NED has changed its original conclusions as to the 
identity of IC 4948 and now makes it equivalent with NGC 6902,  Simbad meanwhile 
continues to equate the identity IC 4948 with the 17th mag galaxy ESO 284-18.
 - Malcolm J. Thomson

As with all corrections to the NGC and IC Catalogues, there is a certain margin
for error, even though the evidence supporting the correction may be strong and
compelling. It is with this in mind that we ask the user to use this information
as 'Most Probable', but never to assume the correction is 'Absolute'.  All
published corrections are based on an exhaustive 'paper chase' of the historical
record back to the original discoverer's published notes/papers, and are
therefore based upon the historical accuracy (or inaccuracy) of those particular
notes/papers. In short, Caveat Emptor! - Robert E. Erdmann, Jr.