In Seaman's Practice (1637), the English mathematician, diver, navigator and surveyor Richard Norwood described how to estimate the speed of a ship at sea, using a line knotted at intervals of 50 feet and a 30–second sandglass. Each knot that went past a particular point in the 30 seconds represented one nautical mile per hour; if (say) ten knots go past, the ship is travelling at ten knots, or ten nautical miles per hour.
The nautical mile represents the length of an arc on the Earth's surface, along a line of longitude, corresponding to one minute, or one sixtieth of a degree. In other words, it's a defined fraction (1 divided by 60 x 360) of the circumference of the Earth, measured along a line of longitude. In practice, it's rounded (in the UK) to 6,080 feet.
Fifty feet in 30 seconds is equivalent to 6,000 feet per hour; Norwood's method is thus a close approximation – to within one and one third per cent (approximately 1.33%) of the true value.
The international (metric) standard for the nautical mile is 1,852 metres (6,076 feet). For this standard, the distance between the knots is about 14.4 metres (approximately 47 feet 3 inches).
Norwood's method used his own calculation for the length of a line of longitude, which he made by measuring the distance between London and York and observing the sun's altitude in the two places. His result was some 600 yards too great; but it was the nearest approximation that had been made in England up to that time, and was noted by Isaac Newton in his Principia Mathematica (1687).
What I'm not sure about is what's used, at sea, as the stationary point against which the knots are counted.
© Haydn Thompson 2022