Before the fourth race at Pimlico one day last week, a familiar type of announcement came over the public address system: "In the sixth race, scratch No. 1, I'm Jama."

The horse in question was part of a stable entry. Both I'm Jama and No. 1A, Ineluctability, were trained by Francis Campitelli. Under the rules of Maryland racing, Campitelli could wait until the fifth race was official to decide whether he would run one or both horses and to make final the riding assignments on those horses.

Of course, anyone who was wagering on the Pick Three (which comprises Races 4, 5 and 6) and thought he was betting on I'm Jama was out of luck. Anyone who placed advanced wagers was out of luck. Anyone who had handicapped the race and figured that I'm Jama's speed would be a factor in the way the race developed had to go back to the drawing board. And anyone who tried to change a wager he already had already probably was out of luck too.

"Basically, we have a very strict policy about cancellations," said mutuels director Elizabeth Quill.

Surely, there were bettors at Pimlico and at the Laurel simulcasting who probably never knew I'm Jama was out of the race at all. Public address announcements in the middle of a racing card often go unheard. When an entry is part of a race, half the people at the track will be asking the other half: "Did you hear any changes?" I have been known to bet a horse and watch the field cross the wire before I realized that my horse wasn't even in the race.

Why are trainers allowed to make late changes to an entry when they wouldn't be allowed to scratch a single horse? Steward Clinton Pitts says this is the customary scenario: A trainer has entered two horses in a race on Tuesday, but there may be a race on Thursday for which one would be eligible.

However, entries won't be taken until midafternoon on Tuesday, and not until then will the trainer know for certain if the Thursday race is going to be run. If not, he'll keep both horses in the Tuesday race. If so, he'll scratch one to run on Thursday.

"The feeling of horsemen is that the system is good the way it is now," Pitts said. Of course they feel that way. Only the public is inconvenienced by the present system.

However, one member of the Maryland Racing Commission feels that the state's rules governing entries deserve an overhaul. Jack Mosner says he will propose some changes at the board's September meeting.

Mosner would require trainers to name their jockeys and make scratches at the normal scratch time -- the afternoon before the race -- even if they are entering two horses in a race. But he would eliminate most entries by permitting horses with the same trainer but different owners to run uncoupled.

Because I'm Jama and Ineluctability have different owners, they would race as separate betting interests. The trainer would therefore be in a position to bet on one of his own horses while betting against another.

"That's the way we do it now in stakes races," Mosner argued. "I feel that it's contradictory to have a different set of rules for overnight races. Besides, people want full fields, and this would give us more betting interests in a race."

As much as most bettors would love to do away with the annoyances caused by stable entries, this is not the way to do it. Mosner either has a short memory or else he doesn't realize what happened when this very experiment was once tried in Maryland.

Uncoupled entries created such negative fan reaction that the commission was forced to go back to the old system (which has existed for most of the history of the sport.)

Whenever a trainer had two uncoupled horses in a race, bettors' suspicions were kindled. Which one was he trying with? And when the longer-priced part of an entry prevailed, they were irate.

Pitts recalled: "I was at Pimlico when Sonny Hine ran two horses in a race; one was 4 to 5, the other one was 45 to 1. When the favorite was up the track, and the 45-to-1 shot won, they were ready to burn the track down."

Moreover, the argument that was advanced at that time -- that split entries would create large fields -- proved dead wrong. Split entries enabled the racing secretary's office to card small fields.

If five horses were entered in a race -- and six were needed to put the race on -- the secretary could say to one of the five trainers: "Give me another horse and the race will go." There were races in Maryland in which a six-horse field would have two horses trained by Bud Delp and two trained by Dick Dutrow. Talk about potential for fan paranoia!

But the wrongness of Mosner's main idea should not deter the commission from examining the policies governing stable entries and stopping the 11th-hour scratches and jockey changes that confuse and annoy fans.

Scratches and jockey changes in entries should not be permitted after a day's racing program has begun. The interests of the betting public ought to take precedence over the desires of a small number of trainers.