Okay — twenty-five. Not a pinch, not a scoop. Twenty-five, count them, they're lettuce, they're free. Beans you do ten and you'll feel it, beans you can hear. Fold the flap over twice and run your thumbnail down it or it opens in the box on the way to the fairgrounds. Three hundred forty envelopes before the 4th and it's quarter to eleven.

Line one is the lot letter, top right, big. Line two common name, line three whatever Latin the donor wrote down — you copy it even if it's wrong, you don't fix it, that's not your job today. Line four is the year it was collected, not the year it came in. Line five germ percent and then the test date right next to it, same line, no exceptions. A percent with no date on it is not a number, it's a rumour. Half this rack came in as rumours.

It's eighty-eight in here. It was eighty-eight most of August. You want the rule, here's the rule: degrees plus humidity under a hundred. Fifty and forty-five, fine. Seventy-five and sixty, you've got a packet that will not sprout in two years. We're at eighty-eight and thirty-one and I've asked for a window unit and a dehumidifier three budget cycles running.

And the moisture matters about ten times what the heat does, so when somebody tells you they put their seed in the icebox, ask them if they dried it first, because if they didn't they've done more harm than the shelf would have. Dry and warm beats cool and damp. Every time. Shatter test — fingernail, if it dents it's not dry, if it snaps it's dry.

Sort them by how fast they die, not by the alphabet. Onion and parsnip are done in a year and nothing you do saves them. Lettuce and pepper, three, four. Tomato, cucumber, the cabbage family, five to eight. A melon seed dried right will still come up at ten years. Anybody who files this rack A to Z throws out half their onions every year and never knows why.

Count of good lots — I had forty-four. Forty-four. No. Forty-three. I counted the '88 parsnip, lot AE, and the '88 parsnip is dead, and I know it's dead, and I wrote it down anyway. Parsnip is one year. It was dead in 1990.

Forty-three. Forty-three. Forty-one. Lot M and lot R are both squash, both from D.V.B., both off Ashlan, and squash is bees, and they cannot both be what they say they are. One of them is the other one or they're both something in the middle. That's where exchanges actually go wrong — not rot, not bugs. Four hundred feet between a zucchini and a crookneck and you get seed that looks perfect and comes up as neither, and nobody finds out till July, in somebody else's garden, and they blame themselves.

Corn's wind, so a quarter mile or you bag the tassels. Beans, peas, lettuce, tomato — those pollinate themselves, they need next to nothing, you can grow four kinds of bean in one bed. Selfer or outcrosser is the only sort that matters and it is not printed on one packet in this rack.

Forty-one. Forty-one. Forty. I just read eighty-eight percent off lot AB and there is no test date on lot AB. Caught myself. Eighty-eight percent of when. Doesn't count. Don't distribute it and don't write it on a fresh envelope.

Anything under about sixty, you don't thin it, you sow it double and you stop putting it out. That's not a judgement on the donor, that's arithmetic.

Marisol — here. Lot Q. Cilantro, '89, from J.K.T. Eighty-two percent tested April of '90, which is a real number with a real date on it, which is more than I can say for most of what's in front of you. I've never grown it out. Fill twenty of those.

Tomato you ferment, three days in a jar of its own juice on the sill, it takes the gel sac off and kills most of the canker. Seed scraped straight off the cutting board keeps both. Dry-seeded stuff is threshing and winnowing and nothing else. And don't pick the big ones out — big seed out of the same lot is not better seed, you're just selecting for size and against everything else that mattered.

Hybrid seed comes in here about twice a year, in good faith, every year. F1 tomato seed doesn't give you the F1. It gives you a scatter of the parents and most of the scatter is worse.

Eighty-eight degrees. Memo went in 8/14/91, that's the second one, first was March. Window unit and a dehumidifier. The propagation room has the towels and the propagation room is seventy-one because it has a vent that works.

The one thing I'll say for myself: every packet I've labelled with my own hand since 1984 has a test date on it. Every one. That's two thousand six hundred and forty envelopes and not one of them is a rumour.