On several of the land masses I detected scattered deposits of the soil-type I needed, and I could have replenished the _Uxurient's_ soil-solution vat from any of them. However, I chose an unusually rich one on a large island near the major land mass, reasoning that the less time I consumed in the operation, the less chance there would be of my occasioning "a substantial interference in the normal evolution of an extra-Community culture". This particular deposit bordered a small community of scattered, thatch-roofed dwellings, and abounded in trees similar to the yumquat species. After activating the _Uxurient's_ ventral camouflage-unit, I brought the greenship down to about two hundred mirids, gravved it into position above the edge of the forest, and opened the capillary-tube lock. I timed my maneuver to coincide with the passing of the dusk belt, but, reluctant to attract any more attention than was absolutely necessary, I waited through most of the ensuing night phase before lowering the capillary tube. Unfortunately, I erred somewhat in my calculations, and the tube's rhizomorphous feeding system, owing partially to the rather strong wind that had sprung up during the night phase, entered the soil much closer to one of the native dwellings than I had intended should be the case; however, dawn being near at hand, I lacked sufficient time to recoil and relocate the device, so I left it where it was. I was not particularly worried: the natives' superstitious fear of the tube would probably preclude their approaching it closely enough for them to be able to damage it, and if their superstitious fear of the tube itself was not strong enough to make them keep their distance, their fear of the "low-lying cloud" from which the tube depended should be. My mind at ease in this respect then, I reduced the opacity of the hull's upper hemisphere to complete transparency so that the greendeck would benefit from the rays of the system's sun, after which I retired to the subdeck to check on the first influx of nutrients into the soil-solution vat. The length of the capillary tube prohibited any immediate change in the solution-level, so while I waited, I busied myself checking the tubes that run down to the vat from the section of the greendeck where the upper extremities of the capillary tube are affixed. Next, I checked the outgoing tubes that feed the greendeck soil. By the time I finished, the level had begun to rise. I waited till it rose above the halfway mark, then I took a sample and ran an analysis. The result delighted me: the D-2 and Z-1 mineral element content had quadrupled! If the rapidity with which the vat was filling continued, I would be able to disengage the capillary tube, recoil it, and be on my way before the next night phase.