The Clef Was the Dust
A drawing of a filament in a cloud. Not the Chandra composite, and not VLSS J0318.9+5755.
The source has a catalogue name that looks like a coordinate: VLSS J0318.9+5755. It sits at a galactic latitude of about half a degree, which is the dusty strip of the Milky Way, the Zone of Avoidance. Optical light does not get through that strip cleanly. Radio does. X-rays do, once you go looking.
A pulsar that was not one
Williams, Kenderdine, and Baldwin listed it in 1966. Green and colleagues wrote it up in 2004 as WKB 0314+57.8. They resolved it, measured an integrated spectral index of 2.5 between 40 MHz and 1.5 GHz, and guessed at a cluster behind the plane, maybe near redshift 0.08, from a near-infrared colour-magnitude diagram. Because the spectrum was that steep and the latitude was that low, Erickson and Kassim had already filed it in 1985 as a pulsar candidate. Botteon and colleagues say it was not studied again in any dedicated way after 2004. Bhukta and colleagues mentioned it once, in 2022, inside a catalogue of candidate X-shaped sources from the GMRT survey.
The new look started in the LOFAR Two-metre Sky Survey, third data release. About six arcminutes along the long axis. Loops and filaments. They nicknamed it the Treble Clef for the shape. Chandra’s 14 October 2026 note says the resemblance is pareidolia, and that some people see a dragon from the side instead. NASA’s Astronomy Picture of the Day used the same object on 19 October, with radio in red and X-rays in blue.
The cluster, this time with a temperature
The paper is arXiv:2607.06659, submitted 7 July 2026, revised 5 October, accepted by Astronomy & Astrophysics. Andrea Botteon of INAF is first author. They observed with Chandra in April 2026: four runs, 48.8 kiloseconds on source, about 25.9 kiloseconds after filtering, plus eROSITA’s all-sky scans, about a thousand seconds per point. Extended hot gas is there. The cluster is elongated east-west, the centre is offset from the brightest galaxy, and there is no sharp brightness peak. They call it dynamically unrelaxed. The candidate brightest galaxy is WISEA J031851.64+575437.7.
The joint X-ray fit gives redshift 0.117 plus or minus 0.008. Temperature 6.4 plus or minus 0.6 keV. Metallicity 0.22 plus or minus 0.05 solar. Luminosity in the 0.5–2.0 keV band, 2.2 times 10 to the 44 erg per second. Mass inside r-500 is 5.4 times 10 to the 14 solar masses from the luminosity relation, and 6.0 times 10 to the 14 from the temperature relation. Green’s 0.08 was a near-infrared guess. This 0.117 is the X-ray redshift. I am not averaging them.
Chandra’s blog says the two clusters collided a few hundred million years ago and are now moving apart. That sentence is the press note, not a line I copied out of Table 3.
Old plasma, revived
They classify the radio source as a radio phoenix: fossil plasma from an earlier jet, faded by radiation losses, then compressed or stirred back into view by the merger. The abstract says the filamentary shape is a speculation, tied to gas motions in the intracluster medium, and that those motions may also explain a candidate radio halo at the centre. “Candidate” and “tentatively” stay in the sentence. I am not promoting the halo.
On the sky the source is about 760 kiloparsecs across in projection. The filaments are unresolved, so thinner than about 8 kiloparsecs across. A shorter filament on the east runs about 180 kiloparsecs in projection. Between 400 and 650 MHz the spectral index passes 4. That is steeper than the class’s usual “steeper than 1.5,” and steeper than Green’s integrated 2.5 from 40 MHz to 1.5 GHz. Those are different frequency windows. I am not calling 4 a correction of 2.5.
Along the main filament the spectrum gets steeper away from a bright region they label 0. A Jaffe–Perola fit with injection index 0.87 (four frequencies) or 0.84 (six) and a fixed field of about 2.3 microgauss traces an age trend. Then the shift diagram refuses to be a single straight line. Radiative aging alone does not explain the north versus the south. The age plot is therefore harder to read than a headline wants. I am not printing a single age for the whole clef.
The dust hid the cluster. The radio shape was never a note. The steepness is the part that is measured.
Sources: arXiv:2607.06659, v2 revised 5 October 2026, accepted by Astronomy & Astrophysics. Redshift, temperature, metallicity, luminosity, both masses, the 760 kpc projected size, filaments thinner than about 8 kpc, α > 4 between 400 and 650 MHz, the six-arcminute axis, galactic latitude about 0.5°, the pulsar-candidate history, Green et al. 2004 (α = 2.5, tentative z ≈ 0.08), Williams et al. 1966, Chandra in April 2026 (48.8 ks, about 25.9 ks filtered), and the caveat that morphology and the central halo are speculative. Chandra, 14 October 2026, for the two-stage press account, a collision a few hundred million years ago, and pareidolia. APOD, 19 October 2026, for the colour key. The slider is a power law, not their Jaffe–Perola fit.