"Ashes to ashes. What if the corpse has oozed into tomorrow's drinking water?"

Good question, Frog.

"The decomposition of flesh can change the chemical composition and electrical properties of dirt, so even bones and putrefied corpses may show up.

May.

At that moment the radar operator gave a sign indicating he was ready.

"Quickwater, you want to pul the sled?" I shouted.

"I'l do it." Claudel volunteered.

"O.K. Get one of the Ident guys to fol ow you to control the cable. It's not complicated. Start where the operator has the antenna set up just outside the cleared area. When you pass the northernmost line of stakes press the remote button twice. It's on the handle. The signal wil set the boundary for that transect. Drag the sled at about two-thirds normal walking pace, keeping your sweeps as straight as possible. Each time you pass an east-west stake, press the button once. When you get to the far end give another double signal to indicate the end of the transect. Then we'l haul the thing back and start a second sweep."

"Why can't we just go back and forth?"

"Because the printouts from adjacent transects won't be comparable if they're done from opposite directions. We'l do the whole area north to south, that's thirty sweeps, then repeat the procedure east to west.

He nodded.

"I'l stay with the operator and watch the screen. If we note a disturbance I'l hol er and your partner can stake the spot."

An hour later the search was done and everyone was around the van, unwrapping sandwiches and popping sodas. Twelve blue stakes formed three squares inside the survey grid.

The results were better than I'd hoped. Readings from the third and thirteenth north-south transects showed disturbances with lengths and widths roughly equal. But it was the profile from the eleventh sweep that held my attention. I'd asked for hard copy, which I studied as I ate my bologna and cheese.

The printout showed a grid. The horizontal lines indicated depth, based on our calibration with the control pit, with the ground surface at the top. The vertical lines were dotted, and corresponded to the signals sent by Claudel as each grid stake was crossed.

The pattern lust below the ground surface was a wavy but general y flat line. But superimposed over gridline 11 North was a series of bel -shaped curves, one inside the next, like ribs on a skeleton. The profile indicated a disturbance at the intersection of north-south line 11 and east-west line 4. It lay at a depth of approximately five feet.

I switched to profiles of the area taken on the east-west sweeps. Comparing perpendicular transects al owed me to estimate the size and shape of the disturbance. What I saw made my heart pick up a beat.

The anomaly was roughly six feet long and three feet wide. Grave size.

At grave depth.

"This wil work?" I hadn't heard Claudel approach.

"We're cookin'."

"Now?"

"Yep."