Neither rain, nor snow, nor heat, nor gloom of night stays these couriers from the swift completion of their appointed rounds.

That's what the legend says. But what if the stamps fall off the letters? What if the mailbox leaks in winter storms? What if a postal clerk sits on a chair and it breaks?

The answers to these and other perplexing challenges to swift mail delivery are hunted daily by the 201 employes at the U.S. Postal Service engineering support center just south of Rockville. Engineer Hassell Crouch, for instance, was at work one recent afternoon trying to find out if a letter box was strong enough to earn the postmaster general's stamp of approval.

With the help of a rod and a pulley, he connected the letter box door to a little electric engine that opened and closed it 7,500 times in three hours -- pretty much what the life of the average letter box boils down to, he explained. The task of raising and lowering the little red flag lay ahead.

"Once in a while you get one with an aluminum hinge and it won't take it," said engineer Tom Lanyi, who was also working on the testing. "And sometimes the hardware falls off the flag."

The experiment was one of many conducted daily at the facility in Montgomery County, where lab workers say they receive a regular supply of letter box samples and prototypes. The samples must pass official inspection before the product can be sold to the public.

In another building, Les Wall, who is in charge of "quality assurance" at the Post Office labs, suspended an aluminum letter box in a large green metal vault connected to a maze of hoses and dials. He opened the heavy lid, and thick salty steam billowed upward. After 72 hours, he said, he would have a good idea of how well the letter box resisted corrosion.

Other standard tests for letter boxes include pouring more than 15 quarts of sand over them to see if the paint holds up, pouring water over them to see if they leak, setting fire to plastic ones to see if they are fire resistant and putting loads of up to 200 pounds on them to see if they bend or break.

"We look at a box for general appearance," Crouch said, "and sometimes we might recommend rejection of a box if it is too attractive to somebody stealing it. But you are kind of walking a tight line there."

But just as important as the inspections for letter boxes is the steady flow of tests the lab conducts on equipment used within the Post Office. After all, employes point out, the operation is a major corporation with operating expenses this year projected at $24.1 billion and contracts worth more than $1.5 billion.

A few feet away from Wall's salty steam chamber, new labels for mail sacks were basking in brilliant artificial sunlight as they revolved slowly inside a big machine called a "Weather-Ometer." A flick of a button could add driving rain and fierce winds to the test of durability.

And besides the millions of dollars' worth of sophisticated machinery used in the research, the Post Office uses a lot of ordinary items.

Frank Thurston, who heads the office of engineering services at the Rockville labs, pointed out a cage-like device used to test the strength of chairs. It raises them about six feet in the air, then drops them seat-first on a small platform. "A lot of people want to sell chairs to the Post Office," he said to explain why the Post Office checks the chair strength.

Other testing equipment includes a "tumbler," a giant mechanized version of a hamster-cage treadmill used to test the strength of parcels and packages, giant showers that imitate rainstorms, as well as much of the machinery found in normal postal sorting centers, which engineers are constantly tinkering with and trying to improve.

But as the mail sorting and delivery systems have become increasingly mechanized, the Post Office's research has become more sophisticated. Although much of the testing equipment "is not sophisticated at all," said Warren Denise, the center's acting director, "analysis using computer systems is much more sophisticated, and used much more."

The world of postal research has changed much since the center was established in 1950 with a staff of three people. Among the first devices that were researched, tested and deployed throughout the country were a twine-tying machine and a machine to pick up hampers of mail and dump them on conveyor belts.

By the mid-1960s, according to engineer Charles Shaw, who has studied Post Office research history, engineers were working on large letter-sorting machines. By the mid-1970s, work had begun on optical character readers, which would electronically read a ZIP code.

So while mailboxes were being opened and closed in one part of the lab, engineer Joseph Peng was peering through an electron microscope at a tiny sliver of paper cut from an envelope.

Peng explained that he was looking for traces of phosphors on the envelope with the 20-cent stamp printed on the surface. The ordinarily invisible phosphors, magnified 200,000 times, appeared on the microscope screen like lumps in tapioca pudding.

All stamps these days are coated with phosphorus material, and the ink used in automatic franking machines contains small amounts of the substance, which glows under an ultraviolet light but is normally invisible. Automatic sorting machines can detect the phosphorus material and line up an envelope so the address can be read and the stamp can be canceled.

Peng said workers at the Post Office labs are constantly monitoring the phosphors in stamps and ink. Too little or too much means it will not be detected by the machines.

Other recent experiments by Peng included an analysis of a conveyor belt to find its weaknesses, analysis of a metal die used in stamp cancellation to see why it was breaking, and microscopic and X-ray analysis of a Post Office 35-year service pin to see if it contained as much gold as it was supposed to -- which it did.

While such work may seem trivial, postal officials say it is not. Gerald F. Merna, who manages the huge Northern Virginia postal center in Merrifield, said that anything that can make the 1.5 billion pieces of mail his center handles each year flow more smoothly makes a big difference.

During a recent evening mail-sorting rush, he pointed out a crew of seven people using a large machine to sort 9,600 large envelopes each hour into 100 containers for different destinations.

The box-like containers used to be on the floor, Merna said, and lifting them was slow, back-straining work. Engineers from the postal laboratories modified the machinery so the boxes were waist-high and slanted toward the machinery so the mail would not bounce out. The mail is moving more quickly, he said, and there are fewer injuries.

Back at the Rockville labs, engineer Tom Lanyi was using dummy stamps to experiment with a new stamp-dispensing machine that he had built into a counter, like ticket machines at the movies. But he said it was not as easy to build as the machines at the movies.

First of all, he said, dehumidifiers and heaters had to be installed so the stamps would not stick together. The machine had to dispense several different denominations of stamps and cut them off precisely on the perforated line. And it had to fling them out of the slot far enough that the clerk would not have to touch them -- but not so far they would fall over the edge of the counter.

Once all these things are worked out, Lanyi said, he hopes his machine will find its way into post offices everywhere. Not only will they save the clerk and the customers time, he said, but also, "It will be more hygienic."