You have twelve nights on the 100-inch. Each night you swing the telescope onto one nebula and expose a plate. At the eyepiece you set a comparison star beside its brightest star and bring it nearer or push it farther until your eye is satisfied. At the desk you put the spectrum under the measuring microscope and turn the screw until the crosshairs sit where you say they belong. Then you pencil one dot on your chart. Bright nebulae take a night; faint ones, two or three.
Hubble's distances were about seven times too short, we now know, and today's rate is about 70 km/s for every million parsecs. The relation itself held. He wrote that it "may represent the de Sitter effect", and left the obvious consequences for later.
The sky. The smudges are nebulae, the faint far-off ones beyond our own stars. Tap one, and the card below tells you its name and what it costs: bright ones one night, faint ones two or three. You have twelve nights, the lamps along the top. Then tap Point the telescope.
The eyepiece. Through the eyepiece the nebula breaks into stars; its brightest is ringed in amber. The brightest star of any such nebula is about as bright as any other's, in truth: it looks faint only because it is far. On the right half of the field sits a comparison star just like it, which the dial brings nearer (it brightens and grows) or pushes farther away (it fades and shrinks). Put the comparison star where it looks as bright as the nebula's brightest star, by your own eye; the readout says how far away you have put it, and that is your reading of the nebula's distance. Tap Read the distance.
The spectrograph. A sandwich of three strips. Above and below, the light of a calcium lamp in the laboratory, with its two amber lines: the lamp does not move, so that is where calcium's lines sit at rest. Between them, the nebula's light, with the same two calcium lines as dark bars. If the nebula is moving away, its bars sit to the right of the amber lines (redder); if it is coming toward us, to the left. Turn the measuring screw: the two crosshairs start on the amber lines and travel across; the readout is how far they have travelled, as a speed. Put the crosshairs on the dark bars, by your own eye, and tap Read the speed.
The chart. Each nebula becomes a dot: distance across, speed up. Red is going away, blue is coming toward us. After your dot lands, a faint hollow dot shows where Hubble's own table puts that nebula. Watch the shape the dots make, and choose your next nebula accordingly.
The Sun. When the nights are over, one tap takes out the Sun's own motion through space, which Hubble worked out at the same time. The dots move.
The rate. Lay the ruler through your dots. Flat means the speeds do not depend on distance. This is your answer; commit when it fits. Then say how fast a nebula far beyond your chart is going, and watch its spectrum come in.
Twelve nights on the 100-inch at Mount Wilson. Each night you point at one of Hubble's twenty-four nebulae, read its distance in the eyepiece from how faint its brightest star looks, read its speed in the spectrograph from how far the calcium lines have slid, and drop a dot on the plate. Then the Sun's own motion comes off, you set the rate, and Humason's far nebula tests it.
In January 1929 Edwin Hubble put together the twenty-four nebulae whose distances he trusted (from Cepheid variables for the nearest, from the brightness of their brightest stars for most, from the average brightness of nebulae for the four in the Virgo cluster) and the radial velocities that had been measured for them, mostly by Slipher and Humason. The velocities had been a puzzle: a "K term" of several hundred km/s that would not sit still, and a solar motion that came out differently each time. Hubble let the velocity grow with distance and solved for the Sun's motion in the same step. The result: the velocity rises by about 500 km/s for every million parsecs of distance (465 from the nebulae taken singly, 513 from nine groups), with residuals of about 150 km/s, which he read as the nebulae's own motions. Twenty-two more nebulae without distances, run through the relation backwards, came out with the same spread of true brightness as the first twenty-four, which he took as independent support. Humason's new velocity for NGC 7619, 3779 km/s, fitted. Hubble called the relation "a first approximation representing a restricted range in distance" and raised the possibility that it was "the de Sitter effect", a prediction of relativistic cosmology. He did not use the word expansion.
This game demonstrates the pattern in Hubble's Table 1. The twenty-four nebulae on the sky carry the table's distances and velocities; the eyepiece runs the paper's brightest-star criterion (brightest stars at a true brightness of −6.3 magnitudes) and the spectrograph runs the ordinary shift of the calcium H and K lines; the Sun's motion is taken out with the paper's own solution for it. The check that ships with the game re-solves Hubble's four-unknown equation on the table and recovers his constant, his solar motion and his residuals.
It assumes the game's framing. The twelve nights and the one-, two- and three-night costs are ours: Hubble's programme ran for years, and he says only that the far nebulae are the hard ones. The eyepiece reads every nebula by the brightest-star rule, including the six Hubble did by Cepheids and the four Virgo nebulae he did by cluster brightness; where the table gives no brightest-star magnitude the rule is run backwards from the table's distance. The nebulae's positions on the sky are from standard catalogues, not the paper, and are spread apart so they can be tapped. Your readings are your own; the instruments add no error beyond your matching, and the glow that says you are close is ours. The Sun's motion is applied from Hubble's solution, not solved by you. Humason's NGC 7619 is placed at "about seven million parsecs", the cluster distance the paper quotes as independent of the velocity.
The paper establishes a roughly linear relation between distance and radial velocity among these nebulae, with a constant of about 500 km/s per million parsecs on its distances, and a solar motion of about 300 km/s. It does not establish what the relation means; Hubble explicitly left that open. It does not establish the constant we use today: his distances were about seven times too short (the Cepheid scale was wrong and some of his "brightest stars" were clouds of glowing gas), and the modern value is about 70 km/s per million parsecs. The plate at the end redraws his Figure 1 from his table, with his line. Your dots are your readings of his numbers, not new evidence.
Table 2 (the twenty-two nebulae with velocities only, and the matching distributions of brightness that Hubble offered as support); the nine-group solution; Lundmark's earlier attempt; the remark on galactic rotation; everything about de Sitter beyond one line on the end screen.
Edwin Hubble, "A relation between distance and radial velocity among extra-galactic nebulae", Proceedings of the National Academy of Sciences 15 (1929): 168–173. doi:10.1073/pnas.15.3.168. Open access at pnas.org.
Game built from this paper; the paper is the authority where the two disagree. Fonts: Jost and IBM Plex Mono (SIL Open Font License). The game number in the page address fixes the look of the plates and star fields; the sky and its numbers are the paper's every time.