beyond the overlap point. A loss of shunt or other cause of false pick up of the track relay 
with a train in section could not, therefore, cause the repeater relay armature to rise. No other 
railway, however, adopted this system, due to R. P. Brousson and A. H. Binton, (Patent No. 13709 
of 1904). The tubes connected with the District Railway had the same equipment as used on that 
line, except for small modifications necessitated by the more restricted clearances. The signals 
on the District line were lighted by gas, with Adlake long burning lamps as a stand-by, but those 
in the tubes by long burning lamps. These arrangements gave place to electric lights later and to 
light signals proper in due course. The same track circuit system was adopted by the Metropolitan 
Railway between Praed Street Junction and Aldgate in 1909 and between Baker Street and Neasden in 
1911, after which all new work used AC track circuits. J. B. Struble had patented the vane relay 
in the United Kingdom (Patent No. 2899 of 1904) and a complete AC installation had been brought 
into use on the New York Subway in October of that year, following a smaller trial one in 
California, marking the beginning of very important developments. The last polarised DC relay was 
not removed, however, from the London Underground running lines until 17th November 1951, at 
Whitechapel. 

Although all signals were fitted with train stops it was thought advisable to have some 
additional equipment on the open sections to meet fog conditions, and a number of electro-
pneumatically driven Clayton type detonator placers were installed, maintaining a fog signal on 
the rail a short distance in rear of a signal at danger. The magazine held 50 fog signals and on 
the number falling to a certain minimum an air whistle began sounding, warning the man in charge 
of the machines. This working was superseded eventually by the fog repeater signal - an additional 
lamp signal at driver's eye level - first tried on the Metropolitan line during the 1914 war, in 
consequence of the great difficulty of providing fogmen on the open sections of line, and which 
gave extremely good results. 

It will, of course, be realised that a great deal had to be learnt by practical experience in 
day to day working before the best results could be obtained with the new and, in some respects, 
novel equipment. One of the greatest difficulties was the lack of a really good insulating 
material able to withstand the particularly onerous conditions obtaining, and satisfactory spring 
contact material. 

Trouble was experienced also with corrosion and expansion of piping, condensation of water in 
pipes, the effects of vibration, and the breaking of insulated joints and bond wires. The 
original single solid bonding had to be replaced by the double flexible type and many small 
changes in the design of certain items effected. The fixings of the train stops had to be 
modified, for instance, after the collision at Caledonian Road, Piccadilly line, on 4th 
September 1912, when a motorman ran at speed past a home signal at danger and struck a train 
broken down in the platform, the train stop failing to engage with his trip valve. Nevertheless 
excellent results were obtained, and no fatal accident fell to be recorded with the original 
apparatus. Improvements were constantly made as experience was gained and the number of failures 
in service - of which, in any case, an infinitesimal number were on the danger side - were 
markedly reduced over the course of time. This pioneer work proved the great value of automatic 
and power signalling for lines such as the District Railway, and its example was followed not 
only in Great Britain but in other countries, as on the Berlin Elevated and Underground line, 
the original equipment for which was supplied by Westinghouse.