Many recent authors (e.g. Way 2013, ASPC 471, 97; Peacock 2013, ASPC 471, 3) are critical of the fact that Hubble (1929, PNAS, 15, 168) did not provide any citations for the velocities that he lists in the two tables in his paper. These authors are concerned mainly with the first purpose served by citations, feeling that Hubble received credit for "discovering the expansion of the universe" to the exclusion of others equally or more deserving. Without engaging in argument, one should also point out that Hubble is not the first to omit such references - Young & Harper (1916, JRASC, 10, 134) and Wirtz (1922, AN, 215, 349) are predecessors who failed to do so, and de Sitter (1930, BAN, 5, 157) would be equally negligent in a subsequent paper.
What these authors neglect to point out, however, is that the second purpose for citations is also not being served - in this case the ability to track back data to their source. Is there any reason to care? After all, the number of velocities involved is small, and most were measured by Slipher, whose early publications are now well indexed.
In this case, yes. Big time. The actual velocity of NGC 1700 is +3900 km/s. All the values used by the above authors are completely bogus! What happened? Surely the papers that do provide proper citations can help us trace the origin of these values. Right?
Not exactly. It turns out that the practice of providing citations at this time was more haphazard than one would imagine. Sometimes a reference is provided to the name of an author, but not the publication. Sometimes a reference is provided to the publication, but not name of the author. Many times a comment would be made that Slipher was the source of most of the velocities, but the source of the rest would be left unmentioned. Another common practice is to provide the names of all authors whose values are cited in a table, but no attempt is made to sort out which value goes with which author.
Thankfully one author does a good job. Stromberg (1925, ApJ, 61, 353) collects together all the velocities used by Lundmark, Lemaitre, and Hubble in one table, and he cites the origin of each velocity. [The fact that Stromberg doesn't actually list the publications for Slipher's velocities is of no consequence here.] For NGC 1700, Stromberg gives the value +800 km/s, and for the source he gives a reference to Pease (1918, PASP, 30, 255). Ahah! Pease must be the culprit who originated this misinformation. Right?
No. Here is what he said:
"The spectrogram is too much under-exposed to admit of trustworthy results, but a rough measurement reveals a high positive radial velocity."
That's it - no velocity at all! The only person to cite Pease properly is Eddington (1923, Math. Theory Rel.):
"... NGC 1700 has been observed by Pease, who found a large receding velocity but gave no numerical estimate."
Stromberg, however, never even bothered to read Pease! What is going on? Here is what I have found.
The problem with this practice is that the invented velocities can take on a life of their own. The value of +500 is repeated by Wirtz (1922) in a table of 29 velocities (for which he does not give any reference.)
It is in Lundmark's 1924 paper that we find the first appearance of +800 km/s, again without reference. This value is also reported in Stromberg (1925) (notwithstanding that Stromberg cites Pease 1918 for the source) and repeated in Hubble (1929). Where does this value come from?
Here is my guess. It turns out that R. F. Sanford is the one who measured a velocity of +800 km/s from the spectrum obtained by Pease, but it was reported, not in the Pease paper, but rather in the Annual Report of the Director of Mt. Wilson (which may be found in the Carnegie Institution of Washington Yearbook for 1918, p. 213). Further, the velocity is marked as being uncertain. Does anyone actually cite the Annual Report or indicate that the velocity is uncertain? Not that I can tell.
In a final fit of madness, de Sitter (1930) reports a value of +700 km/s. I have no explanation for where that value came from. (Perhaps he found one +500 and two +800 values in the literature and averaged?)
In summary, the literature citations for NGC 1700 have proven to be essentially useless. No one pays them the least attention. One may as well have left them out. They give credit where none is due or desired.
Pease (1920, ApJ, 51, 276) repeats his assertion regarding NGC 1700 that "... the nebula has large positive radial velocity" but, pointedly, does not give a value.
NGC 1700 is not the only velocity to be reported incorrectly in an observatory report. The velocity of NGC 4884 (a Coma cluster galaxy) is given as +1500 in the Mt. Wilson Observatory Report published in Popular Astronomy, 37, 70 (1929). In Humason (1931, ApJ, 74, 35) the correct value of +6700 is given. In a footnote, Humason gives the explanation for the first erroneous value. Basically, one should never trust a redshift based on the detection of a single line.
The reason for tracking down the velocity of NGC 1700 in the first place is that it is the worst outlier in Lundmark's "Hubble" diagram. Peacock dismisses this object along with seven other outliers as being "dwarfs" (only two are), but NGC 1700 is one object for which Lundmark found the right distance - it is a bad velocity that makes it stand out.