Bidston Kelvin Machine

Most of the funding for the establishment of the LTI in 1919 had come from Sir Alfred Booth and his brother Charles Booth of the Booth Shipping Line in order to “prosecute continuously scientific research into all aspects of knowledge of the tides” (Doodson, 1924; Carlsson-Hyslop, 2020).

In 1925, Charles Booth also provided most of the GBP 1500 required for the purchase of a 25- to 28- component TPM for the exclusive use of Doodson at the LTI. The BAAS contributed GBP300. In the event, this became a 26-component machine, later extended to 29 components. This Bidston Kelvin Machine (TPM-S14) was made by the firm of Kelvin, Bottomley and Baird of Glasgow with Doodson taking a great interest in its construction and suggesting modifications. It stood about 7 feet (2.1 m) high, 6 feet (2 m) long and 2 feet (0.6 m) wide and was encased by sliding doors (Doodson, 1925, 1926, 1927).

Then in 1929, following the death of H. W. T. Roberts (the son of Edward Roberts), Doodson acquired the Légé-made machine that Edward Roberts had designed in 1906 and which had won a Grand Prix at the Franco–British Exhibition of 1908. Roberts had subsequently used it as part of his own tidal prediction business (Messrs. Edward Roberts & Sons of Broadstairs), with the work of that business, including pre- dictions for the Hydrographic Office, also passing to the LTI in 1929. Doodson paid the Roberts family GBP 753 15s 0d. This Roberts–Légé Machine (TPM-S5) simulated 33 con- stituents, several more than the Bidston Kelvin Machine. By 1929 it was in need of an overhaul and refurbishment. Shortly thereafter, the number of constituents was increased to 40 by Chadburns of Liverpool, with its original design having allowed for such a future expansion. Its dimensions are approximately 7 feet (2.1 m) high, 6 feet (2 m) long and 2.5 feet (0.8 m) wide (somewhat wider with its casing). It re- mained in use at Bidston until 1960 (Scoffield, 2006). This TPM was sometimes known as the “Universal Tide Predic- tor of 1906” and also as the “Roberts Tide Predicting Ma- chine”, a name which had previously been attached to the In- dia Office Machine (TPM-S2) which had also been designed by Roberts. Doodson and Bidston staff referred to it as the “Légé machine”.

Doodson wrote in an internal note that there was no difference in performance between the Bidston Kelvin and Roberts–Légé machines. He stated that for both “the total error of production of a tide does not differ by more than 0.05 ft (1.5 cm) and 1 min of time from calculations using the same harmonic constants, even for the largest ranges of tide”.

These two TPMs equipped Doodson with experience that would prove invaluable when considering the design of later machines. They would also provide him with the means to enable the LTI to become one of the main producers of tide tables worldwide. The two machines were to play particularly important roles in World War II, in providing tidal predictions for D-Day and for other military operations in Europe and the Pacific. Those for D-Day were unusual in that they were made for an unknown location (now of course known to be Normandy), using only the crudest of harmonic constants provided by Commander Farquharson at the Tidal Branch of the Admiralty (Parker, 2011). However, they were not secret for long. In July 1944, Doodson gave a lengthy interview to the Liverpool Daily Post, stressing the importance of the TPMs to the D-Day predictions (LDP, 1944).

In 1947, Doodson asked Légé to estimate the price of a machine with at least 35 components. Légé produced an initial design which Doodson started working on in detail. This became the 42-component machine delivered finally by Légé in December 1950 at a cost of GBP 5049 (Woodworth, 2016).6This Bidston Doodson–Légé Machine (TPM-S20), referred to by Bidston staff as the “D–L machine”, was the third machine to be used operationally at the LTI and one of a number of similar machines made by Légé & Co. under Doodson’s supervision and exported to several countries. Doodson (1951) describes it in some detail. It was a single-sided TPM, being converted to a double-sided machine in 1956. It has an aluminium frame, unlike the steel frame of TPM-S5. The machine is approximately 7 feet (2.1 m) high, 9.5 feet (3 m) long and 3.5 feet (1.1m) wide and weighs 1.4 t (not including the support base frame). Complete engineering drawings and documentation survive which have enabled its recent restoration.7Figure 3f and g provides photographs of the machine.

The purchase of the Doodson–Légé Machine had been made contingent on the sale of the Bidston Kelvin Machine to the Service Hydrographique et Océanographique de la Marine (SHOM) in France. It was acquired by them in 1950 and was operated first in Paris and then in Brest for predicting the tides of overseas ports (SHOM, 2019). It remained in operational use at SHOM until 1966 (Rawsthorne, 2019).

Nowadays, the Roberts–Légé and Doodson–Légé machines are both on display at the National Oceanography Centre in Liverpool (Tide and Time, 2019), the only place in the world where two TPMs can be seen alongside each other. Meanwhile, the Bidston Kelvin Machine may still be found at SHOM in Brest. All three machines are in working or near working order.<ref>https://www.researchgate.net/publication/340064506_Tide_prediction_machines_at_the_Liverpool_Tidal_Institute#pf6 Tide prediction machines at the Liverpool Tidal Institute, March 2020History of Geo- and Space Sciences 11(1):15-29, DOI:10.5194/hgss-11-15-2020, Philip L. Woodworth