Researchers allowed horses to watch plastic apples being placed out of sight in buckets. The horses were then found to opt for the bucket holding the most apples.

The researchers, from the University of Essex, used fake apples to ensure horses weren't making the judgement through use of smell.

The findings indicated that not only were horses keeping some kind of measurement of how many apples were going into the buckets, but they were then able to make a call on which to select.

The results placed horses on a par with 10-month-old human infants, rhesus macaque monkeys and lemurs.

Dr Claudia Uller and Jennifer Lewis used 57 untrained horses belonging to local owners and a riding school for their series of experiments.

The animals were first allowed to nibble a small piece of real apple to get them interested. Real apples were then replaced with identical fakes.

The first test involved two plastic apples being placed in one bucket and three in another under the watchful gaze of the horses. The containers were then held at eye levels so the horses could make a choice.

Eleven of 13 horses who underwent the test opted for the bucket with three apples.

The second test involved one bucket holding a single large apple while the other held two smaller ones. Ten of the 12 horses tested chose the two-apple bucket.

Dr Uller, herself a horse rider, told the British Psychological Society's annual meeting in Dublin: "The result absolutely proves that horses are more intelligent than people think."

She argued the ability may have an evolutionary origin, with the option of going for more food possibily being a trait hardwired into the brains of all animals.

"It's a very basic, rudimentary capacity," she said of the horses. "An obvious question is whether they're going for more numbers, or more something else, mass or density."

Dr Uller was inspired to conduct the research by the century-old story of a horse named Clever Hans, who impressed German audiences with his apparent ability to perform arthmetic.

The horse, owned by a German maths teacher, could seemingly add and subtract numbers by stamping on the ground.

It was exposed as a fraud by psychologist Oskar Pfungst in 1911, who found that Clever Hans was in fact responding to unconscious nods and gestures from members of his audience.