With the opening of the through route to the Midland Railway at Olney, between Northampton and Bedford, via the Easton Neston Minerals, Towcester, Roade and Olney Junction Railway (yes, really!) the EWJR was able to offer the facility to the Midland of a short cut for freight traffic from Bristol docks to London. This was particularly useful for the perishable traffic of bananas. Trains arrived from Bristol at Broom Junction, where they were reversed over the EWJR. Initially Midland locos were used, but they were too heavy for the track and the East and West needed to provide its own power. However, the banana trains used stock fitted with the automatic vacuum brake, and the EWJR had no suitable locos. An approach was made to the LNWR, who were selling some of their famed DX goods locos, and three were bought at the end of 1892. One, LNWR 1891 (formerly 648 until 1887), had the vacuum brake fitted and was therefore technically an SDX – S standing for “special”: it became EWJR number 7 . The other two, LNWR 1945 (809 until 1889) and 1966 (825 until 1890) were not fitted with power brakes, and became general freight locomotives numbered 8 and 9. The working timetables had three paths allocated for banana trains, but the existence of the vacuum brake on number 7 only suggests that at first, only one train ran this way.
The DX goods locos bought had been fitted with new fireboxes and boilers under Webb, but these three appear to have had them built to the Ramsbottom pattern and dimensions (shorter firebox, larger boiler at a lower pitch – modellers beware!) which also had a top hinged smokebox door.
One of the last two was fitted with cast-iron wheels with 12 spokes of H-section instead of the usal 15 spoke variety. This was fairly uncommon, and they were nick-named as “duckfoot DXs”. Numbers 7 and 8 had been rebuilt by the LNWR in 1878 and 1880 respectively, but number 9 was not rebuilt, suggesting (to me, at least!) that it was number 9 which was the Duckfoot loco as delivered. In 1903, number 8 was reported as having worn out wheels, and number 9 as being worn out apart from the wheels, so the obvious solution was to put the wheels from number 9 under number 8. Although I have been able to track down a photo of number 8 with duckfoot wheels, I do not have the date, so whether this is after the swap, or before it, I do not know for sure but I suspect the latter as the locomotive has acquired brakes on the loco, not just the tender, and appears to have had a vacuum ejector fitted – something not, so far, documented elsewhere, and which I surmise was added when it was rebuilt. This suggests that number 8 had some substantial work carried out in 1903, to make one working loco out of two that were getting worn. Despite this, it was listed as “recently withdrawn” in 1908. As a modeller, I can choose which version suits me until I find out more, and even then it will require nothing more than a renumbering if I have things wrong as the locos were otherwise identical.
Number 7 was placed onto the duplicate list in 1908, renumbered as 07, and scheduled for scrapping but number 10 was in a worse state and number 7 was renovated and fitted with a new boiler of standard Webb pattern from the LNWR in 1914 and taken back into capital stock as 7, lasting until 1920.
It is somehow typical of the EWJR that they managed to buy three examples of one of the most numerous and famously standardised locomotive classes built in the UK, and all 3 were different!