Arp’s Phantom Jet


During the “Great Debate” of 1920 astronomers Herber Curtis and Harlow Shapley had a famous debate on the nature of “spiral nebulae”. Curtis argued they were “island universes” or what we would today call a galaxy. Shapley was of the opinion that they were spiral structures within our own galaxy. One of the evidences Shapley put forth was that another astronomer, Adriaan vanMaanen, had reported detecting rotation of these objects over a period of years leading to an overall rotation rate of ~105 years. If these spiral nebulae were truly as far (and thus, as large) as Curtis suggested this would mean they would be rotating well beyond the speed of light at their outer edges.

It was later determined vanMaanen’s rotation was a case of wishful thinking when Hubble eventually determined the true distance to the Andromeda galaxy. From then on, it was well established that galaxies are so large, their motions will not be observed in human lifetimes. Aside from local flare ups of supernovae and other such events, galaxies should be relatively static. Yet in just over 40 years, a distinct, large-scale feature in the galaxy NGC 3303 seems to have disappeared entirely.

In 1964, Halton Arp observed NGC 3303. This oddly shaped spiral galaxy he reported as having a jet protruding from the northwest side. It made it into his famous 1966 compilation of photographs entitled, “The Atlas of Peculiar Galaxies” as Arp object 192. A 2006 publication by Jeff Kanipe and Dennis Webb (The Arp Atlas of Peculiar Galaxies: A Chronicle and Observer’s Guide) listed this jet as a “challenge” for astronomers to capture.

In 2009, an advanced amateur named Rick Johnson attempted a long exposure of NGC 3303. When his image was finished, it was notably lacking the jet. The news of this eventually reached Kanipe and Webb and they suspected that the exposure was simply not long enough to have captured this object. To be sure, they consulted images of the galaxy from the Sloan Digital Sky Survey. The jet was missing from these images as well. A major feature on a galaxy had vanished in 45 years and no one had noticed until 2009.

The only plausible explanation was that the jet Arp detected didn’t really exist. It was possible it was a photographic defect in the glass plate on which the image was taken. Another possibility was that the imaged structure did exist, it just wasn’t what Arp suspected.

When Charles Messier attempted to look for comets, he kept a list of 109 objects that were not comets so he wouldn’t be confused by them. To tell true comets apart from the other fuzzy objects he observed, he observed them over a period of nights. If they moved with respect to background stars, they must be relatively nearby. If not, they were likely very distant. Was Arp’s jet the opposite; A nearby object that had simply moved out of the field of view since his original image?

Kanipe contacted the Minor Planet Center to determine if any of the known asteroids or minor planets had been in the vicinity when the image was taken. It turned out that a minor planet, TU240, discovered on 6 October 2002 by the Near Earth Asteroid Telescope on Haleakala, Maui, Hawaii, was very near to NGC 3303 when Arp imaged it confirming it was a strong candidate for Arp’s disappearing jet.

This isn’t the first time an object has been pre-discovered and its true nature simply missed when it was imaged. There is evidence that the planet Neptune was observed at least three different times (including by Galileo) before its nature was understood. But for this TU240,  this is expected to be the earliest prediscovery photograph. As a result, TU240 was given a new designation just after Thanksgiving 2009. It is now listed as 84447 Jeffkanipe.

(Read this story as told by Rick Johnson at the BAUT Forums.)

19 Replies to “Arp’s Phantom Jet”

  1. Awesome! I love it when this sorta thing happens. Mundane explanations can be fun!

    There is evidence that the planet Neptune was observed at least three different times (including by Galileo) before its nature was understood.

    Oooh? Where can I read more about this?

  2. While I know it isn’t the best way to do this, but I looked at the original image on the Caltech site -

    Downloaded the image and messed around a little with it with my photo processing software (Paint Shop Pro).

    I did a high pass sharpen and the “tail” on the end of Arp 192 (NGC 3033) looks slightly bifurcated, when zoomed in.

    This might be nothing more than a jpeg artifact and film grain, but certainly looks odd for an asteroid trail.

  3. Most distressing to read (in Rick Johnson’s account) that the original plates taken by Arp for his atlas are missing! When were these plates last seen? Hopefully, this will spur someone to track down this irreplaceable collection.

  4. There is a very faint feature in both the Arp and SDSS images. It’s a faint line starting very close to the central feature and downwards, slightly to the left. The line is lighter in these negatives and a little more prominent in the SDSS image.

    WriteNoMore, can you post your processed image somewhere for us to see?

  5. Jon Hanford,

    I was wondering if there were plates of the same object taken in the same session. Wouldn’t there be several taken in different colors (red, blue)? If so an astroid would have moved from plate to plate.

  6. WriteNoMore: The bending and splitting at the end of the tail is attributed to field rotation (the rotation axis of the telescope not aligned correctly with the NCP).

    Jon: Photographic plates weren’t often taken in numerous filters. Often they were taken without filters at all. I don’t believe that sequential images were taken at that time.

  7. “”minor planet, TU240, discovered on 6 October 2002″”

    This probably should be minor planet 2002 TU240 discovered on 6 October 2002

  8. Seems Bill Keel wrote a paper back in 1985 that included a spectra (albeit very noisy) of the base of the ‘spike’ taken with the 2.1m Kitt Peak scope. Ironically, the picture published to illustrate the position of spectra taken across NGC 3303 (Arp 192) is non other than Arp’s original from the atlas! The paper, “Recognition and classification of galaxies with optical jets” can be found here:…..90.2207K

    ND, as JV pointed out most of these images were purely ‘shoot-n-scoot’, due to the heavy demand for time on the 200 inch. Only a few would be revisited, days to months later, and alas, Arp 192 was not that lucky 🙂

  9. I guess we’re living in a very fortunate time where repeated surveys such as that from the SDSS and large sampling observations such as from Kepler are possible. So much of astronomy is just brief and incomplete glimpses into the universe.

    Being able to collect and observe large number of objects for extended periods will change so much of our understanding. The only drawback will be the amount of data being generated and storage.

    To quote someone I can’t remember the name of “Too much of a good thing is wonderful!”

  10. Jon: I’m curious as how you can define the “base” of the jet given it was superimposed on the galaxy (if indeed it even existed)!

  11. Whoops, didn’t mean to imply a physical jet or connection. 🙂

    Figure 6 of the paper shows an aperture placement (upper right) near the intersection of the galaxy with the asteroid trail. Figure 7c shows a very noisy spectra labeled NGC 3303 “JET”.

    At the conclusion of the paper, Dr. Keel thanks Arp for the use of the original plates and measuring equipment in Pasadena. So Bill got to see the (now missing) real deal. 🙂

  12. Just thought I’d respond to some of the questions raised here. The apparent bifurcation noted by WriteNoMore, indeed the brightness gradient and curve of the spike, may be due to reciprocity failure. The asteroid is in retrograde, hence it begins over the galaxy when the emulsion was fresh. It “petered out” after 40 minutes (the exposure length) as the emulsion tired. The slight curve may be an illusion caused by the fading track blending nonuniformly with the background.

    I must admit that I, too, was shocked to hear that the Atlas plates are missing. Apparently, however, this is not an isolated incident. Astronomers I’ve talked to say that, in fact, it’s rather common. Over the years plates are inadvertently disseminated by other observers who use them for their research and then fail to return them to the archives, and even by the astronomers who made the plates. Photographic plates, I am told, are actually the property of the observer who made them. I checked with Dr. Arp. He doesn’t have them. He told me they should be in the attic at the Santa Barbara address of the Carnegie Observatories in Pasadena, but they are not there!

    The faint line seen so plainly in the SDSS image, as noted by ND, is actually an absorption feature, probably dust. This is more apparent in the positive SDSS image, seen at the NED site. This very strange galaxy needs to be observed by HST.

    Finally, as to Jon Hanford’s other comment concerning Bill Keel’s 1985 paper, Keel did not image Arp 192 but rather used an aperture plate with a number of 2.5-arcsecond holes drilled at interesting locations, including the jet’s putative position. Since the jet wasn’t there—in fact, didn’t exist—he did not detect emission lines.

  13. I basically agree with Jeff K. on what he said about the emulsion and exposure effects. It would be nice to see the original photo in something like a FITS or other lossless format. It is a shame that originals are now misplaced.

    The other thought that I had (and it is purely speculative) was that the passing asteroid was tumbling (or had a companion) just fast enough for it to get caught on the film.

    Its to easy with a jpeg to get sidetracked by induced artifacts. What I did was just a simple experiment with a common photo processing program, nothing special. The slight curve and widening of the end of the tail in the original photo is was caught my eye.

  14. I took both the Arp Atlas image of 192 and the SDSS image and subjected them to all kinds of contrast-brightness enhancements, filters, and sharpening techniques in an effort to understand how something with the structure of the spike could vanish (this was before I had consulted with the log book of the 200-inch). I found a number of what WriteNoMore says were “induced artifacts.” And yes, it is easy to be distracted by these. Ultimately, Occam’s razor made the best and simplest cut toward an explanation of this mystery.

    The best Arp galaxy images we have today are found on NED in the Arp Atlas section. These images are scans of intermediate photographic negative prints that are a few processing steps away from the original glass plates. We’re fortunate to have them to work with.

Comments are closed.