"Now," said Ishie, "Confusion say he who can fly on wings of mosquito fly better on wings of eagle. How much thrust do we want, Mike?" "What are our limits?" asked the practical engineer. "Limits, schlimits. We got _power_. Of course," he added, "we _are_ limited by the acceptable stress limits on the wheel, and ... yes ... by the stress limits on our plastic, too." "The wheel was designed to stand upwards of 1.5 gee maximum spin--but that's only radial strength," Mike began figuring. "Don't think anybody ever calculated the stress of pulling the hub loose, endwise. No reason to, you know, and it wasn't expected to land or anything. And really, nobody expected it to stand in service more than a 1.5 gee spin on the rim. They computed these racks to take all kinds of shock, but the overall structure is rather flimsily built." He paused for thought. "We could maybe put a tenth of a gee on the axis, but I better check some of the stress figures against the structural pattern with the Cow first. We'll have to give some thought to strengthening things later, if we really want to go into the fantastic possibility of landing this monster anywhere." Consulted, the Sacred Cow computed a potential maximum stress-safety at the hub of something over two-tenths of a gee, and the two finally settled on one-tenth as well within the limits. "Now the other limit," said Ishie. "This little piece of plastic will only stand a pressure approaching the point at which it begins to distort and run out of the field. This stuff is quoted to have a compression-yield strength of one hundred ten pounds to the square inch. We probably shouldn't exceed ... hm-m-m ... ninety pounds. Let's get the Cow to tell us how big a chunk of surface area that represents." The answer was discouraging. Mike rapidly converted the figure in centimeters to feet, and came up with nearly an eighty-three foot diameter for a circular surface. "Looks like we'll have to put it out on the spokes," he muttered in disgust, but Ishie shook his head quickly. "No need, Mike. Later on we'll need a few thrust points out on the rim for good aiming, but we don't have to have all this surface area in one unit or even in one place. Also, we do not need to consider only the surface of an homogeneous piece of plastic material. "This plastic can be cast. Very easily. In it, we can insert structures that will absorb the strain from many surfaces within, rather than only on a front surface. "I expect some of the glass thread with which the hull of the ship was made could be inserted with no trouble. Each thread, then, would take up the strain, and a mass of them distributed through the plastic could deliver a greatly increased amount of thrust from a volume of plastic rather than from a surface area." * * * * * Mike started to object. "To get an absolutely parallel magnetic field, the gap between the pole faces can't be very wide."