He ran his hands back through the hair above his ears, concern momentarily doubling the mass of creases on his face. «It's the timing of the thing,» he said eventually. «Penny wanted us to make a bid for independence as soon as the cloudscoop was operational. Six to eight weeks from now.»

I let out a soft whistle. «That soon?» That wasn't in Zimmels's briefing. I'd gathered the impression they were thinking in terms of a much longer timescale.

«Penny wanted that date because that way she'd still be alive to see it happen. Who can blame her?»

«But you didn't agree.»

«No, I didn't.» He said it almost as a challenge to me. «It's too soon. There's some logic behind it, admittedly. With an operating cloudscoop we can guarantee uninterrupted deliveries of He3 to Earth. It's a much more reliable system than sending the shuttles down to pick up fuel from the aerostats. Jupiter's atmosphere is not a benign environment; we lose at least a couple of shuttles each year, and the aerostats take a real pounding. But the cloudscoop—hell, there are virtually no moving parts. Once it's functioning it'll last for a century, with only minimal maintenance. And we have now established the production systems to keep on building new cloudscoops. So when it comes to He3 acquisition technology we're completely self-sufficient, we don't have to rely on Earth or the O'Neill Halo for anything.»

«And biotechnology habitats are also autonomous,» I observed. «You don't need spare parts for them either.»

«True. But it's not quite that simple. For all its size and cost and technology, the JSKP operation here is still very much a pioneering venture; roughly equivalent to the aircraft industry between the last century's two World Wars. We're at the propeller-driven monoplane stage.»

«That's hard to credit.»

«You've talked to Pieter Zernov, I believe. He's full of dreams of what the habitats can eventually evolve into. We need money for that, money and time. Admittedly not much in comparison to the cost of a cloudscoop; but nor is it a trivial sum. Then there's Callisto. At this moment I've got a team there surveying the equator for a suitable mass driver site. JSKP is planning to start construction in 2094, and use it to fire tanks of He3 at Earth's L3 point. There will be a whole string of tanks stretching right across the solar system. It'll take three years for them to arrive at L3, but once they start, delivery will be continuous. A mass driver will eliminate the need for ships like the Ithilien to make powered runs every month.»

«So what are you worried about? That Earth won't supply the parts for a mass driver? They'll be acting against their own interest. Besides, you'll always find one company willing to oblige.»

«It's not the availability of technology. It's the cost. The next decade is going to see JSKP investment in Jupiter triple if not quadruple. And it's only after that, when there are several cloudscoops operational, and the mass driver is flinging He3 at Earth on a regular basis, when you'll start to see the cash flow reversing. Once we've established a He3 delivery operation sophisticated enough to function with minimum maintenance and minimum intervention, the real profits are going to start rolling in. And that's when we can start thinking about buying out the existing shareholders.»