No engine had ever been made that was capable of pushing a comet core around the solar system at any appreciable speed. For that, they had needed to embed the nuke-on-a-stick into the heart of the ice payload, construct an ice nozzle behind it, and then pull out the control blades, causing the reactor’s sixteen hundred fuel rods to become very hot. Ice turned to water, then steam, which shot out the nozzle and produced an amount of thrust actually capable of making a difference. So a few months had then been consumed disassembling Ymir and integrating its parts into a chunk of ice carved off the three-kilometer ball.

The question might have been asked: Why just a piece of it? Why not bring the whole comet core back, if water was so desirable? What was the point of sending a large nuclear reactor into space if you weren’t going to use it? And the answer lay in the fact that even a large nuclear reactor did not even come close to having enough power to move such a big piece of ice. The mission would have lasted more than a century, assuming the existence of some kind of a miracle reactor that could operate at full power for that long. In order to get this done in any reasonable amount of time, they could only bring back the bare minimum of ice needed to rendezvous with Izzy and accomplish the Big Ride.

In any case, Sean and his surviving band had used the nuclear engine to impart a delta vee of about 1,000 m/s to the shard they had carved off Greg’s Skeleton, thereby placing it into a somewhat different orbit that had, a few months later, glided into L1. Sean had remained alive just long enough to yank out the control blades one last time and execute a delta vee that had basically reversed the maneuver they’d used to leave the L1 gate almost two years earlier. This had simultaneously brought Ymir into geocentric orbit while executing, as cheaply as possible, the plane change needed to enable a later rendezvous with Izzy. A couple of days later Sean had tapped out the “coming in hot, high, and heavy” message and dropped dead. Of what, they could only conjecture.

The retrieval team that was now being organized by Markus was going to use a MIV, or Modular Improvised Vehicle, assembled from a kit of parts: a sort of Lego set for the construction of spaceships, neatly sorted on a stack of modules, collectively known as the Shipyard, connected to the Caboose.

The Shipyard was a generally T-shaped contraption. One arm of the T’s crossbar, projecting from the port side of the Caboose, was studded with MIV parts. The opposite arm was a cluster of spherical tanks surrounding a collection of splitters. These used electrical power to split water molecules into hydrogen and oxygen, and piped them to chillers, which refrigerated the gases until they became cryogenic liquids that could be stored in the bulging tanks.

So much for the T’s crossbar. Its long vertical stroke was a truss terminated by a nuclear reactor: not a small RTG like the ones on the arklets, but a true reactor, originally designed to power a submarine, considerably souped up for this task.

Markus dubbed the Shipyard’s first product New Caird, after a small boat that had been used in Shackleton’s expedition to Antarctica. She was assembled and made ready for use in ten days: about one-third of the time they estimated it would take for Ymir to arc in from L1 and make her closest pass to Earth.

To design, assemble, and test such a vehicle so quickly would have been unthinkable two years ago. During the interval between Zero and the White Sky, however, the engineering staffs of several earthbound space agencies and private space companies had foreseen the future need to jury-rig space vehicles from standard parts such as arklet hulls and existing rocket engines, and had provided a kit of parts, lists of procedures, and some basic designs that could be adapted to serve particular needs. In effect, New Caird had been designed a year ago by a large team of engineers on the ground, all but three of whom were now dead. Those three had been sent up to join the General Population. Building on their predecessors’ work, they were able to produce a general design—enough to begin pulling the bits together, anyway—within a few hours of Markus’s decision. Details emerged from their CAD systems as they were needed over the following week and a half, and the necessary parts and modules were shuttled about the Shipyard until the new vehicle was ready.

New Caird would have to execute one burn to reach an orbit that would intersect Ymir’s and another to match her velocity, so that the crew could board the ghost ship and take the helm. The total “mission delta vee” for that journey, from its departure from the docking port on Izzy to its arrival at a similar docking port on Ymir, was some 8,000 meters per second.

The conversation turned now to mass ratio: a figure second only to delta vee in its importance to space mission planning. It simply meant how much propellant the vehicle needed at the start of the journey in order to effect all the required delta vees.

Laypersons tended to substitute “fuel” or “gas” for “propellant,” making the obvious analogy to the stuff that had been burned by the engines of cars and airplanes. It wasn’t a bad analogy, but it was incomplete. In addition to fuel, most rocket engines needed some kind of oxygen-rich chemical (ideally, just pure oxygen) with which to burn it. Cars and planes had simply used air. Rockets stored the oxidizer in a separate tank from the fuel until the moment of use. The two chemicals were collectively referred to as “propellant,” and their combined weight and volume tended to dominate space vehicle design in a way that hadn’t been true of, say, automobiles, whose gas tanks had been small compared to their overall size.

A convenient figure for characterizing that was the mass ratio, which was how much the vehicle weighed at the beginning (including the propellant) divided by how much it weighed at the end, when all the tanks had been emptied. If you knew how good the engine was, and how much delta vee you needed, then the mass ratio could be calculated using a simple formula named after the Russian scientist Tsiolkovskii, who was credited with having worked it out. It was an exponential: a fact that explained almost everything about the economics and technology of spaceflight. For if you found yourself on the wrong side of that exponential equation, you were completely screwed.

When the relevant numbers for the Ymir retrieval mission were jacked into the Tsiolkovskii equation, the result was a mass ratio of about seven, meaning that for every kilogram of stuff—Markus, Dinah, other personnel, miscellaneous robots, etc.—that they wanted to arrive safely at the docking port of Ymir, they needed to allow for six kilograms of propellant at the moment of departure from Izzy. This wasn’t all that difficult to achieve, especially for a vehicle that would never be exposed to the rigors of passage through the atmosphere.

The payload in this case was a single arklet hull that had been augmented with a “side” door: an airlock that could accommodate one person in a space suit. Other than that, it had been stripped to the minimum complement of equipment needed to keep a crew of four alive for a few days. To its mass, of course, needed to be added that of the actual humans and their food and other essentials. The lightness of a bare arklet hull was startling; the newer hulls, made of overwrapped composites, weighed in at eighty kilograms. Stripped of everything that made it comfortable and inhabitable over the long term, and including the “side door,” the maneuvering thrusters, and a reasonable supply of thruster propellant, the mass of New Caird was about ten times that. The humans weighed three hundred kilograms. The rocket motor that would be doing all the important burns weighed another two thousand. So, in round numbers, the payload mass—the stuff that actually had to get delivered to the docking port of Ymir—was some thirty-five hundred kilograms. The mass ratio of seven meant that its propellant load, at the beginning, was going to be some twenty-one thousand kilograms of liquid hydrogen and liquid oxygen.