Family Group Sheet
Family Group Sheet
NameKenwrick Cecil COX
Birth1879
Death1943
FatherRev Cecil Walker COX (1844-1936)
MotherLouisa Florence BRIDGES (1846-1898)
Marriage1905
SpouseCaroline Crichton “Carrie” CHRISTIE
Birth1873
Death1949
Children
Birth1915
Death1988
Birth1909
Death1941
Birth1909
Death2000
Birth1915
Death1989
Birth1907
Death1998
Notes for Kenwrick Cecil COX
Cox arrived in Bamfield [ Canada] about 1902 as 'electrician' probably just after the Cable station was finished (the earliest passenger record is from Jan 1903, but in others he states he was in Canada from 1902). In October 1902 the first 'round-the-world' cable message was transmitted through this station, among many others. This was known as the 'All-Red Line' because it encompassed the British Empire of the time. Six months later, the American cable system sent its first message to Hawaii.

By 1909 (or 1908 according to one passenger record), Cox had moved to another station at Fanning Island near the equator and around 1915 was stationed at Norfolk Island, South Pacific, then in 1922 to Auckland, New Zealand. He returned to Bamfield about 1929 as station manager and retired to England in 1933. During this time he made a number of trips to England to arrange manufacture of various cable machinery, such as the Cox Selenium Magnifier and was granted a number of related patents.

It was in 1901 that Marconi sent his first radio message across the Atlantic. By the 1930's, more economical radio messages were helping to put some of the underseas cable companies out of business. He must have been a talented 'electrician' because it is said that after Cox returned to England, Senor Marconi gave him work. His children remembered having a television in 1937 London. Cox did war-related work during WWII at General Electric in London, where he collapsed and died after an accident at his bench in 1943.


K C Cox, the Pacific Cable and the selenium cell magnifier.
by David Curry

ABSTRACT
The Pacific Cable, conceived as a strategic telegraph asset for the British Empire, laid in 1902, had major transmission problems due to its length that would not be fully solved until the advent of loaded cables. K C Cox succeeded on his own to develop the apparatus necessary to automate the transmission process but, while he was succeeding, his design was overtaken by other developments.

1. INTRODUCTION

In 1902 K C Cox, trained in the telegraph technology of the nineteenth century, went to work for the Pacific Cable Board, and set out to develop his ideas for automating the inefficient manual relaying of submarine cable signal traffic. Subsequent Cox family accounts, a book of recollections by Pacific Cable operators and the internet web site of a British Columbian museum make reference to Cox and his important apparatus, yet histories of submarine cable telegraphy make no mention of him.

The aim of this initial study has been to explore Cox’s career and some of his inventions, in particular his elegant design for a telegraph magnifier, in the context of the Pacific Cable and to understand its significance in the history of submarine cable telegraphy. The intent has been to lay the basis for further research into this corner of early twentieth century telegraphy.
The available space is sufficient only to describe the main findings to date, but it is hoped that it will stimulate interest in an aspect of telegraphy that has seen little research.

2. K C COX (1879 – 1943)

As a boy, Kenwrick Cecil Cox showed an intense interest in practical engineering, frequenting local factories and, when given a watchmaker’s lathe, teaching himself the arts of fine machine work.
In 1896, at the age of sixteen, Cox joined the Eastern Telegraph Company at the Porthcurno cable station, where he received both operating and engineering training. While there, he made full use of the workshop facilities, building himself a lathe that subsequently went with him around the world (Figure 1), and with another mechanician, John Jeffrey, devising a tape perforator, applying for a patent. In 1900 Cox was posted to Vigo as a mechanician, and in 1902 to Gibraltar as an electrician.
At that time, the world-wide network of submarine cables, much of which was owned by the Eastern and Associated Telegraph Companies, was dependent upon manual relaying at each cable station: a long route might involve ten or more such relays. Unlike land lines, the dispersion caused by the cable capacitance and resistance so slurred the signals that individual pulses became indistinguishable after several hundred miles. In addition, the practicality of transmitting the cable code signals charged the cable, causing their effective zero to wander. The consequence was that the signals received from one cable could not be manipulated by electro- mechanical relays for relaying to the next.

Figure 1. K C Cox at his lathe

However, that cable code, with positive pulses for dots and negative for dashes, made it possible still for operators to read the individual letters, using the signal shapes, and this they could achieve at speeds of 200 letters-per-minute or more. The “translation” process involved reading the signal trace from a siphon recorder and punching a fresh tape for retransmission: it was laborious, expensive and a serious limitation to the capacity of the telegraph circuits.
PDF Creator - PDF4Free v2.0 http://www.pdf4free.com

With first hand experience of these translation issues, Cox was evolving his ideas of how the telegraph apparatus could be developed.

3. THE PACIFIC CABLE


That there should be a submarine cable from Canada to Australasia, as a strategic asset for the British Empire, was originally proposed in 1879.

After much political, commercial and technical argument, agreement was eventually reached by the governments of Britain, Canada, Australia and New Zealand and the cable laid in 1902. A management Board had been appointed to represent those governments’ interests, initial funding provided, and the Board charged with making the cable pay its way; but it is questionable whether the organisation was adequate to manage a major asset that proved to be pushing the bounds of technology. In competition with the Eastern Telegraph and, subsequently, the beamed wireless services, the Pacific Cable was never financially successful: eventually in 1929 it, together with all other British international telegraphic assets, was taken into what was to become Cable & Wireless Ltd.

Figure 2. The Pacific Cable

The chosen route (Figure 2) was: Bamfield (Vancouver Island) – Fanning Island (now Tabuaeran, in the Kiribati group) – Suva (Fiji) – Norfolk Island. From there it branched to Southport (for Sydney) and Doubtless Bay (for Auckland). Agreement was reached with the CPR for use of their land lines to Halifax, and space was leased on one of the trans-Atlantic cables from Britain.

Of all the technical scepticisms, the one that continually dogged the Cable was over its length, particularly for the two northern spans, which were respectively about 3,500 and 2,000 n.miles. The first was 1,500 n.miles longer than anything laid before, and was the longest telegraph cable span ever laid; the resultant, extreme signal distortion limited speeds to about 120 letters-per-minute and the interfering earth currents from geomagnetic effects at times disrupted operations.
The southern part of the system was subsequently improved with direct cables: Sydney – Auckland (1912) and Auckland – Suva (1923). But, not until 1926 was the bottleneck of the northern cables solved by duplicating those spans, using the newly developed, continuously loaded cable. That provided a capability of over 1,000 letters-per-minute, but full advantage could not be taken of that without automatic translation.

4. COX, THE PACIFIC CABLE AND AUTOMATIC TRANSLATION

In 1902 Cox, aged twenty-three, resigned from the Eastern Telegraph and joined the newly completed Pacific Cable, as Station Electrician at Bamfield. From then until 1919, Cox was working successively at Bamfield, Fanning Island and, as Officer-in-Charge, Norfolk Island.
It is not known whether he had been taken on specifically to develop apparatus, but that is what he did, much of it in his spare time, developing a complete system for automatic translation of telegraph signals. The system was working operationally on Norfolk Island from 1917. Cox obtained patent protection for much of his system and also received generous financial recognition from the Board.
PDF Creator - PDF4Free v2.0 http://www.pdf4free.com
In Cox’s translation scheme (Figure 3), the significant stages are the magnifier, relays and interpolator. By magnifying the signal (explained below), it was possible to employ robust moving coil relays: these he developed for the purpose. The magnifier and relay combination corrected for the wandering zero and provided a squared waveform for driving the interpolator. The interpolator took each positive and negative excursion of the signal and broke it up into the appropriate number of pulses. The regenerated signal was then suitable for perforating a new tape, or could be transmitted directly.
His was not the first interpolator: about 1902, S G Brown had introduced a “drum relay” and interpolator, apparently successfully tested on cables of 1,000 n.miles, without the use of a magnifier, although exceptionally delicate in use. Investigation as to why it was not used on the southern Pacific cables, prior to the Cox system, might shed useful light on both.

By 1921, Cox was posted to Auckland and his system was being installed at all the cable stations in the southern sector. Reports to the Board extolled its effectiveness, particularly the major savings in operator costs, but this still left the bottleneck of the northern cables. It was not till 1928, after installation of the loaded cables, that the Board appear to make the decision to install Cox’s system in the north; but, suddenly there was a shift in policy and a completely different system installed.
This was “synchronous regeneration”, a system then also being installed on the Eastern Telegraph’s circuits: it combined translation with an ability to multiplex signals and would be an important factor in developing the world-wide telegraph systems. Cox magnifiers had by then been installed at the northern stations: but it is not known if these were used in the new system. The fitting of the Cox translation system in the south had actually predated by about three years that of synchronous regeneration on the Eastern Telegraph’s circuits; why had the decision not made earlier for a full fit of Cox’s system?

Four months after the new system had been successfully demonstrated, Cox was posted to manage the Bamfield station for his final period of service. Soon after, in 1929 the Pacific Cable was taken into Imperial & International Communications Ltd, forerunner of Cable & Wireless Ltd. Cox retired in 1933: at the reception given for his farewell, amongst all the eulogies appreciating his service to the Pacific Cable Board, it was observed that Mr Cox was “a noted inventor of submarine telegraph apparatus, the Cox Magnifier having been used in cable service around the world.”

5. TELEGRAPH MAGNIFIERS

Various attempts were made in the early twentieth century to develop highly sensitive telegraph relays that could handle the distorted waveforms, most noted being the Brown relay, referred to above. However, real progress could not be made until means were developed to magnify the signal currents, of the order of micro-amps. The challenge was to find some physical effect to link, with a high magnification factor, between the cable circuit and a secondary circuit driving robust mechanisms.

A major reason for the successful operation of long submarine cables was the Kelvin siphon recorder. The key was the moving coil sensor driving a low inertia stylus system. Several successful magnifier designs were introduced during the first two decades from 1900, each following such principles, but designed to vary circuit resistances, usually to be detected in a bridge circuit.
The electrolytic magnifier moved a platinum electrode between two plates in a bath of weak acid, to give varying, differential resistances. The Orling jet magnifier had a salt water jet playing onto an inclined plate and deflected by a quartz rod to alter its length, and hence its electrical resistance.

PDF Creator - PDF4Free v2.0 http://www.pdf4free.com
Probably the most successful, but most delicate, was E S Heurtley’s hot wire magnifier: two resistance wires, heated by electric current, were cooled by air streams, such that movement of the wires caused one to heat up and the other to be cooled, resulting in differential resistance changes.
A breakthrough in design would be to dispense with such mechanical linkages, which could be extremely delicate and must have been a nightmare to set up and maintain. That is what Cox achieved, going back to Kelvin’s mirror galvanometer, using light beams and a selenium cell.
6. THE COX MAGNIFIER
The sensitivity of selenium to light was first announced in 1873 by Willoughby Smith, the effect of increasing illumination being to increase its electrical conductance. This had been discovered by accident during experiments on submarine cables. So far as submarine cables went, it ended there, but the effect became very popular with practical inventors, since light sensitive “selenium cells” were easily constructed, giving a tenfold or greater variation in conductance as the illumination was altered. Various industrial and laboratory applications for this useful effect were developed and it is surprising that it was not developed sooner as the basis for a submarine cable magnifier.
In 1914 Cox applied for two patents, covering the application of, and then the design details for constructing selenium cells “especially adapted for use in connection with telegraphy”. By 1916, a magnifier was in operational use at Norfolk Island and a third patent application made, giving the detailed design of the magnifier. Subsequent manufacture was undertaken by H W Sullivan Ltd
The arrangement (Fig.5) was for a collimated beam of light to be split by a slit screen into a series of bars, reflected off a mirror galvanometer onto a selenium cell. This cell (Figure 6) is a series of long, very thin strip elements of selenium, separated by brass foils. The light bars match the spacing of the selenium elements, each straddling a pair of elements, whose resistances thus change differentially as the light is deflected. A multiplicity of light bars and selenium elements multiplied the effect; connection, through the brass foils, into a W heatstone bridge, was used for actuating the succeeding, moving coil relay stages.
Figure 4. Cox selenium cell magnifier
Selenium cell
Mirror galvanometer
Optics
Split screen
Lamp
Figure 5 Cox selenium cell magnifier
PDF Creator - PDF4Free v2.0 http://www.pdf4free.com
The cell was constructed by stacking and clamping mica sheets with interleaving brass foils. The foils, being slightly wider than the mica, left a series of shallow spaces into which heated selenium was melted. Once solid, the surface was rubbed down to an even finish and the selenium annealed to achieve useful, light sensitive resistance.
Across the bridge is a relay for driving the interpolator. Cox had developed a design of robust moving coil relay with sprung contacts that could be adjusted to give a dead band for coping with interference. This relay also incorporated a siphon recorder. To correct for the wandering zero effect, a similar relay, detecting excessive signal deviations, actuated a slugged adjustment of the bridge balance.
Figure 6. Prototype Cox selenium cell
Selenium displays a lagging characteristic as the illumination is reduced. Cox corrected for this effect by adding shunt inductance across the bridge. It appears that this also could be used beneficially to help correct for the effects of signal dispersion.
Cox had achieved an effective magnifier, giving a high factor of magnification, together with the good frequency response of the moving coil on its bifilar suspension, with the zero inertia of the light beam. A contemporary text-book published by the Eastern Telegraph Company for its employees describes very straightforward setting up and maintenance routines; it states that the magnification factor was high enough for the magnifier to be able to drive crude relays, and that it was suitable for use on loaded cables. The present author has estimated the magnification to be of the order of 40,000 to 1, but how this compares with the other magnifiers is not clear. That the magnifier has its own section in this particular book implies probable fitting also in systems other than the Pacific.
Although magnifiers would eventually be superseded by the thermionic valve amplifiers already being developed, there was one last, the capacitance magnifier, a hybrid of electro-mechanical and valve technology. A moving coil drove a rigidly connected vane as the common plate in a differential capacitor in the grid circuits of a pair of triode rectifiers, connected in a bridge. A valve oscillator drove the grids through the capacitor. This design, maintaining the good matching between moving coil and cable circuit, was used as the initial stage in the synchronous regeneration system: it was highly successful and continued to be used up to the mid-twentieth century. This magnifier, rather than the Cox design, was possibly the one used in that final equipment fit on the Pacific Cable.
7. CONCLUSION
K C Cox’s career spanned that of the Pacific Cable Board, and a time of major development in the apparatus for submarine cable stations. At the start, the technology for operating long cables was firmly embedded in the nineteenth century: but in setting out to develop it, he was still using nineteenth century techniques.
Working on isolated Pacific islands, He provided the Pacific Cable with an effective system for automatic translation and relaying. He and his apparatus were much admired by the operators and the Board, as his efforts transformed the system operability. His concept for a signal magnifier was particularly elegant: simple, robust and with a high magnification.
PDF Creator - PDF4Free v2.0 http://www.pdf4free.com
But the pace he could set was no match for the efforts from Europe or the USA, and he was bound to be overtaken. The developing technology was to leapfrog over his work, taking the whole industry forward: Cox’s developments were too late to make a major impact on the telegraph industry.
ACKNOWLEDGEMENTS
The research for this paper would not have been possible without the assistance and kindness of Mary Godwin, Curator and her staff at the Cable & Wireless Archive, Porthcurno, for allowing access to the archived papers and artefacts.
John Liffen and the late Keith Geddes of the Science Museum, London, helped the Cox family ignite a spark of interest some years ago, and John Liffen has recently been of great help to me, in arranging access to the Science Museum collections. Figure 6 is a Cox cell in the Museum’s collection of Cox artefacts (ScM Inv 1976-269).
Assistance from the following libraries and their staffs is much appreciated: Bristol City Library, The British Library, The Institution of Electrical Engineers.
REFERENCES AND NOTES
For this abridged account of the subject, rather than inserting full textual references, the following summarises the main sources used.
Archives & museums (a) Cable & Wireless, Porthcurno: Eastern Telegraph Company, Staff records; Pacific Cable
Board, Staff records, Board minute books. Cox interpolator (probably the only complete Cox equipment in UK); Capacitance magnifiers; Synchronous regenerators.
(b) Science Museum, London: Collection of Cox artefacts, mainly associated with his magnifier (Inv 1976-269); Heurtley hot wire magnifier (Inv 1950-278); Brown drum relay (Inv 1925-171).
(c) The SPARC antique radio museum (British Columbia) states on its internet web-site that it holds many artefacts, including K C Cox items, saved from the Bamfield station, now closed.
Cox family Unpublished papers, including an account of K C Cox by his daughter, the late Mrs I L Curry.

K C Cox patents (GB) (with J Jeffrey) 25,415 (1899); (with J E Dicketts) 789, 790 & 791 (1913); (sole) 399 & 12,361 (1914); 110,374, 121,489 & 124,632 (1916); 165,460 (1918); 161,746 (1920). Cox also took out patents in Australia and USA, but these have not been researched.

History of Cable & Wireless constituent companies K C Baglehole, A century of service, Cable & Wireless Ltd, 1969; H Barty-King, Girdle round the Earth, Heinemann, London, 1979.
Cable telegraphy engineering at end of nineteenth century C Bright, Submarine telegraphs, Crosby Lockwood, London, 1898.

Selenium cells and their applications G P Barnard, The selenium cell, Constable, London, 1930.
Textbook issued by the Eastern Associated Telegraph Companies to their employees J H Stephens (ed.), Textbook on telegraph cable engineering, 3 vols, 1927-28.

Accounts by operators on the Pacific Cable R B Scott, Gentlemen on imperial service, Sono Nis Press, Victoria (BC), 1994.

Contemporary papers and articles The Electrician and Electrical Review A E Foster, et al, The continuously-loaded submarine cable, J.IEE vol.67, pp.475-506, 1929. H Kingsbury, R A Goodman, Methods and equipment in cable telegraphy, J.IEE vol.70, pp477-521, 1932. K L Wood, Empire telegraph communications, J.IEE vol.84, pp.638-61, 1939.
AUTHOR
David Curry may be reached by e-mail at:- david.curry@britishlibrary.net
or at:- Dill House, 69, Corsham, Wilts, SN13 0AS, UK
PDF Creator - PDF4Free v2.0 http://www.pdf4free.com
Last Modified 22 Nov 2014Created 28 Jan 2018 using Reunion for Macintosh