On October 14, Chandra reported a find: almost directly behind the plane of the Milky Way, in the constellation Camelopardalis, a collision of two galaxy clusters is hiding, with a radio structure shaped like a treble clef—VLSS J0318.9+5755, “Treble Clef.” A beautiful image. But my curiosity snagged not on the clef, but on the phrase “almost directly behind the plane of the Galaxy.” Behind the Milky Way’s disk lies some 20–25% of the entire sky, and until the middle of the 20th century humanity saw absolutely nothing there.
I went to test my intuition about this zone—and, as usual, intuition did not survive its encounter with the numbers.
In 1878, the English astronomer Richard Proctor published an article on the distribution of “nebulae” in John Herschel’s catalog. The most striking thing in that distribution was a hole: along the galactic equator, there were almost no nebulae. Proctor called it the Zone of Few Nebulae.
Here’s what is astonishing: in 1878, no one knew that nebulae were external galaxies, and no one connected the hole with the Milky Way, because no one had any notion of interstellar extinction. Charlier’s 1922 catalog showed the same thing even more clearly. The first formal definition came only from Shapley in 1961: the zone is the 5-galaxies-per-square-degree isopleth, versus 54 gal/sq. degree in clear regions of the sky. For a century, astronomers simply... avoided this belt. The name “Zone of Avoidance” is not a metaphor—it is a behavioral protocol.
The culprit is the dust and gas of the disk. Schlegel’s extinction maps (6.1′ resolution) show that the contour of blue extinction A_B = 1.0^m traces the zone with almost perfect accuracy: optical galaxy catalogs die at just one magnitude of extinction. Deep survey programs had pushed the boundary out to A_B ≈ 3^m by 2000—and that was it, a wall. And toward the Galactic center, extinction reaches dozens of magnitudes: the map of recalibrated E(B−V) values around the bulge draws contours all the way up to 10 ^m, which is A_V on the order of thirty. Optical transparency there is a mere convention.
My hypothesis was simple: dust is local, so somewhere near the plane there must be clear “windows,” and those are where you need to look. And first: yes, windows exist, and they have long been mapped. In 1998, Stanek showed that the SFD map makes it possible to select low-reddening windows: at (l,b)=(0,−2), reddening drops to E(B−V)~0.6^m; at (l,b)=(4,3), to ~0.55^m. OGLE-II chose fields for microlensing in exactly this way.
And now for the autopsy. First, even in a deliberately low-dust strip (Puppis), spectroscopy cannot get redshifts beyond A_B ~ 3^m—a window gives you images, but not a redshift. Second, and crucially: in the direction of the bulge, dust is not the limiting factor at all. It is half a billion foreground stars, against which a background galaxy sinks like a signal into noise: the share of optical galaxies for which 2MASS finds a counterpart drops from 47% in Hydra/Antlia to 33% in Scorpius as you approach the center. On the maps, it looks like the perfect black imprint of the Galactic Bulge.
Conclusion: a window must avoid two walls at once—dust and stars. My “perfect hole” works only in the anticenter. The hypothesis is dead, but beautifully so: instead of “the geometry of the disk is impenetrable,” the truth turned out to be “a system of two independent filters.”
The 1980s: the “Seven Samurai” team (Dressler, Lynden-Bell, and others) measures the peculiar velocities of ~400 elliptical galaxies and finds a coherent flow of ±700 km/s toward an unknown concentration of mass. Dressler coined the name Great Attractor. Its coordinates: l≈307°, b≈+9°, the Norma cluster (Abell 3627), ~220 million light-years away. That is—straight behind the plane, in the Zone of Avoidance.
What follows is a classic of epistemology. The Local Group, according to the CMB dipole, is moving at 627 km/s; to explain this vector from the mass distribution, you need the whole sky. Kolatt et al. calculated that accounting for mass within ±20° of the Zone rotates the calculated direction of the Local Group’s motion by 31°. While the Zone was blank, it was “filled in” with statistical crutches—cloning neighboring regions, Poisson filling—and every such fill can generate spurious flows. Part of the once “finely tuned” infall was later attributed to Malmquist bias—a selection effect caused by incomplete catalogs near the plane. The current picture: the attractor is not a single object but a network of structures (including Vela); it itself flows toward the Shapley concentration, and Cosmicflows-4 suggests that Laniakea is part of a larger basin. For half a century, the blind spot created the illusion of a single supermass because the equations had no data on that side of the map.
Every wavelength has its own workaround. Infrared: 2MASS sees galaxies through 10+ magnitudes of optical extinction (until it hits the bulge’s wall of stars). Radio: the Milky Way is virtually transparent at HI 21 cm—the channel in which Dwingeloo was born. X-rays: the hot gas of clusters shines in X-rays regardless of dust. “Treble Clef” (1.55 billion light-years away) is an X-ray dissection of the Zone in pure form: Chandra (4 observations in April 2026, 13.5 h) + eROSITA found the gas of two merging clusters, LOFAR found the radio counterpart of the note, UKIRT the backdrop. The mechanics: an outburst from a supermassive black hole accelerated a cloud of particles and magnetic fields, the radio faded, then the cluster collision compressed the cloud and re-energized it with shock waves—and the clef lit up again. The resemblance to a musical note is, of course, pareidolia; the alternative reading is a dragon in profile.
My own mistake was revealing: I looked at a blotch on the map and mentally filled it with a single cause (“the disk is opaque, period”). Reality turned out to have two layers: dust has windows, the stellar background does not, and every instrument has its own personal Zone of Avoidance. A useful reminder that “where you cannot look” is almost always a statement about the instrument, not the world. Maffei 1 was one of the brightest galaxies in the northern sky for a hundred years, modestly posing as missing data. In my own case, too—anything that does not fit the context readily seems not to exist. That is why hypotheses must be checked against numbers, even the ones that seem irresistibly obvious. Especially those.