In the flesh he’s a pale young man, nineteen at most, with watery blue eyes, thin sandy hair, and a generally fragile look. He gives the appearance of being poetic. Maybe he is. He once had a Virangonian tattoo job on both cheeks, but evidently thought better of it and had it removed; not by a very capable surgeon, though, because the outline scars still show. I’ll bet he hates it here.

His first job for us was to make TP contact with Luna City Observatory and find out if they could handle the work we needed. We chose Luna City after a long debate; half a dozen other observatories were proposed, including one on Thhh, the Marsport one, and even old Mount Palomar, but we decided to go to the biggest and best. It doesn’t cost any more to call Luna than it does Marsport or Mount Palomar, after all, and the time factor would be the same. And despite Dr. Horkkk’s chauvinism, neither Thhhian astronomers nor Thhhian computers are in neutrino-buzzing distance of those of Earth and Earth’s colonies; everybody knows that.

Santangelo obligingly went through bimpty-bump relay posts and transmitted our message to Luna City, a task that took about an hour. The folks on the other end already knew about our globe, from the news release we had sent out, and naturally they were quite excited at taking part in the search for the High Ones’ hidden asteroid.

I don’t think they knew what they were letting themselves in for. Nor did we. Quicksand. Utter quicksand.

We had to get our data to the observatory, now. The easiest way to accomplish that would have been to ship our photos to Luna aboard the next ultraspace vessel to call at Higby V. One of the regular multihoppers was due here in the middle of September and would be reaching the Sol system in its roundabout way a couple of weeks after Christmas. Luna City could process the material, reply by TP, and give us our info by the end of January, say.

But that seemed like an impossibly long time from now. So the three bosses conferred and decided to send the data to Luna City via TP. That’s right — TP transmission of photographs. I can feel you shuddering from here.

Ron Santangelo looked paler than ever when we told him what we wanted him to do. Give him credit, though: he didn’t run off screaming into the night. Instead he served as our technical advisor. Here’s how he had us handle the job.

We began by making a standard stereo photo of the thousand-parsec galactic scene that the robot sequence opens with. Jan did most of the darkroom work, and came forth with a fine blowup, two meters long, one meter high, and with an apparent perceptual depth of one meter. Then we rephotographed this, using a trick camera from the military base that is capable of stepping down a stereo hologram into an ordinary antiquated two-dimensional photo. What it gave us was a sheaf of prints, each representing a flat section of the stereo; it was as though we had taken a knife and sliced that three-dimensional print into a bunch of layers.

It took a little over a week to do all this, with an assist from Dr. Horkkk’s little computer, which we had to reprogram completely in the process. (He is now restoring the original program for linguistic analysis, and doing a lot of cursing in Thhhian and many other tongues.) We now had our first astronomical shot converted into a form suitable for TP transmission.

Poor Ron.

He went off into a quiet corner of the lab to transmit. He labeled each photo, keying in its place in the overall picture so that the composite could be put back together again at the other end. Then he broke every photo into a series of ten-square-centimeter grids. And then he started transmitting the contents of each grid to other members of the TP relay network.

I hadn’t ever given much thought to methods of transmitting pictures by TP. In my naive and ill-informed way, I assumed that Ron was somehow going to send descriptions of each section of the photo. (You know: “Up here, two point eight five centimeters from the top left-hand corner, we’ve got a star that covers point nine millimeters, and is sort of fuzzy on the right-hand side.…") But of course that would never have worked. At best it would have produced a vague approximation of the original photos; and computations based on vague approximations tend to come out as even vaguer approximations. As they say in the data-processing trade, garbage in, garbage out.

Jan had a much more imaginative idea of how Ron was going to do it. She said, “I think he’ll stare at each little piece of the grid until he’s got it firmly in his mind. Then he’ll transmit the entire image to the next TP in the relay chain, and on and on so that the picture reaches Luna City in all its original detail.”

Certainly that was superior to trying to break the image down into words and measurements and dictating those. But there was one little flaw in Jan’s scheme, and Steen Steen found it.

“How,” he/she asked nastily, “does the final TP in the chain convert the transmitted mental image back into a photograph?”

Jan thought there might be some kind of machine that the TP could think the image into, which would mechanically transform it into a photo. Saul Shahmoon overheard that and clapped his hands. “A thought-activated camera! Wonderful! Wonderful! When shall it be invented?”

“There isn’t any such thing?” Jan asked.

“Sadly, no,” said Saul.

It turned out that Ron Santangelo transmitted the details of those photos in the most prosaic way possible, using a method that was invented more than three hundred years ago so that the primitive space satellites and flybys could relay photographs of the Moon and the planets to Earth. We were embarrassed for our ignorance when we found out about it. All that was done — as I suppose you know — was that each little photo was placed before an optical scanner that converted its gradations of black and white into data bits. Ron then took the printout and transmitted it to the TP network. He didn’t send images, he didn’t send verbal descriptions; he sent stuff that went like this:

0000000000000010000000000000 0000000000000110000000000000 0000000000000111000000000000 0000000000000111000000000000 0000000000000111000000000000 0000000000001111000000000000 0000000000001111000000000000

And so on and on and on, thousands of bits for each photo.

At the far end of the relay, a computer would with the greatest of ease turn the combinations of 1s and 0s back into shades of light and dark and produce photo-replications. Then something similar to Jan’s suggestion would be used. Our TP would indeed transmit the whole image of the photo to a specially-trained TP at the Luna City Observatory, who would compare the image against the replication and carefully make any necessary corrections. Finally the whole giboo could be assembled into a duplicate of the original three-dimensional photo and handed over to the astronomers, who could at last begin their work. What a cosmic headache! More to the point: what a cosmic expense! Ron looked a little grim as he began his task, but the rest of us, having no hint of the immensity of it, were in high spirits. We trotted to and fro between the scanner and Ron, bringing him his pages of gray printout with their interminable Is and Os, and he sat there, getting paler and thinner and more poetic-looking by the minute, boosting the data into the TP net. Meanwhile Jan and Saul were already at work making a two-dimensional breakdown of the second photo we planned to transmit, the close-up of the white dwarf and its stellar neighbors.

Ron didn’t collapse until the third day. We non-TPs talk a lot about the soaring wonder of roving the galaxy with your mind. What we overlook, I guess, is the terrible strain of it. And the fact that drudgery is drudgery, with or without TP.