London brewers are among the first potential customers for Watt's improved engines. They have been using carthorses (harnessed to walk round a turning shaft) to work their pumps. So in their case the saving on fuel cannot be the yardstick for assessing Watt's machines.

Instead, these potential customers ask Watt to tell them how many horses can be replaced by his engines.

Watt conducts a series of experiments which lead him to the conclusion that one carthorse, working steadily, can raise 550 lbs one foot higher in every second. This figure (550 ft-lbs per sec) becomes established as the 'horsepower', long used as a basic but in practice rather variable unit of power.

Watt has overestimated the strength of a horse by at least 50% (to his own short-term commercial disadvantage, since his engines can do the work of more horses than he claims). But in the longer term this error is irrelevant. Horsepower never subsequently has any direct link with horses.

London brewers are among the first potential customers for Watt's improved engines. They have been using carthorses (harnessed to walk round a turning shaft) to work their pumps. So in their case the saving on fuel cannot be the yardstick for assessing Watt's machines.

Instead, these potential customers ask Watt to tell them how many horses can be replaced by his engines.

Watt conducts a series of experiments which lead him to the conclusion that one carthorse, working steadily, can raise 550 lbs one foot higher in every second. This figure (550 ft-lbs per sec) becomes established as the 'horsepower', long used as a basic but in practice rather variable unit of power.

Watt has overestimated the strength of a horse by at least 50% (to his own short-term commercial disadvantage, since his engines can do the work of more horses than he claims). But in the longer term this error is irrelevant. Horsepower never subsequently has any direct link with horses.

**Horsepower**

London brewers are among the first potential customers for Watt's improved engines. They have been using carthorses (harnessed to walk round a turning shaft) to work their pumps. So in their case the saving on fuel cannot be the yardstick for assessing Watt's machines.

Instead, these potential customers ask Watt to tell them how many horses can be replaced by his engines.

Watt conducts a series of experiments which lead him to the conclusion that one carthorse, working steadily, can raise 550 lbs one foot higher in every second. This figure (550 ft-lbs per sec) becomes established as the 'horsepower', long used as a basic but in practice rather variable unit of power.

Watt has overestimated the strength of a horse by at least 50% (to his own short-term commercial disadvantage, since his engines can do the work of more horses than he claims). But in the longer term this error is irrelevant. Horsepower never subsequently has any direct link with horses.