But to return to the black gaps. Are they really windows in the star-walls of the universe? Some of them look rather as if they had been made by a shell fired through a luminous target, allowing the eye to range through the hole into the void space beyond. If science is discretely silent about these things, what can the more venturesome and less responsible imagination suggest? Would a huge “runaway sun,” like Arcturus, for instance, make such an opening if it should pass like a projectile through the Milky Way? It is at least a stimulating inquiry. Being probably many thousands of times more massive than the galactic stars, such a stellar missile would not be stopped by them, though its direction of flight might be altered. It would drag the small stars lying close to its course out of their spheres, but the ultimate tendency of its attraction would be to sweep them round in its wake, thus producing rather a star-swarm than a vacancy. Those that were very close to it might be swept away in its rush and become its satellites, careering away with it in its flight into outer space; but those that were farther off, and they would, of course, greatly outnumber the nearer ones, would tend inward from all sides toward the line of flight, as dust and leaves collect behind a speeding motor (though the forces operating would be different), and would fill up the hole, if hole it were. A swarm thus collected should be rounded in outline and bordered with a relatively barren ring from which the stars had been “sucked” away. In a general sense the M8 cluster answers to this description, but even if we undertook to account for its existence by a supposition like the above, the black gaps would remain unexplained, unless one could make a further draft on the imagination and suggest that the stars had been thrown into a vast eddy, or system of eddies, whose vortices appear as dark holes. Only a maelstrom-like motion could keep such a funnel open, for without regard to the impulse derived from the projectile, the proper motions of the stars themselves would tend to fill it. Perhaps some other cause of the whirling motion may be found. As we shall see when we come to the spiral nebulæ, gyratory movements are exceedingly prevalent throughout the universe, and the structure of the Milky Way is everywhere suggestive of them. But this is hazardous sport even for the imagination—to play with _suns_ as if they were but thistle-down in the wind or corks in a mill-race.