I'm enjoying this thread discussing multiple things at once. An interlacing thread, perhaps? I think the Anderson/Fowler/Chambers/Hughes story is rather more subtle that it is often characterised, and all the individuals concerned were real people with strengths and weaknesses. For example, Anderson is often painted as the villain, and as we have discussed he had his blinkers on sometimes. But Cox tells us that the decision to test the Castle on the LMS, and to build a three-cylinder loco rather than four cylinders, and simple not compound, were Anderson's preferences. Which does not fit with the normal way his portrayed, as wedded to Midland Compounds. Perhaps he really did order all those Midland compounds in the early days because the (statistically insignificant) testing they did showed they were the most efficient locos and cheapest to maintain in the Class 4 power bracket, rather than simply because he has stuck in a rut? (Incidentally, Cox' 1946 paper shows that, although coal consumption did increase as they moved to secondary duties, it increased to the level of being similar to simple engines of the same power classification, not worse than them, and the locos were still fairly cheap to maintain compared to other contemporary Class 4 because they were mechanically robust, despite the inside cylinder). Cox also tells us that the excellent 2-6-4T, and the Scots, were essentially designed by Herbert Chambers (in the latter case he was "seconded" to NBL) - previously derided on here as hidebound by Midland bias. But he was responsible for the good front end of both designs. In response to questions from Holcroft and others, Cox stated that the Scots were "pure Fowler" - so Cox at least thought Fowler deserved the credit for these fine engines, as CME and ultimately responsible. Again, not how it is normally portrayed. The same paper interestingly shows the evolution of what became the Horwich Crab, being something like: - Caledonian impressed by performance of ex-Highland River 4-6-0s, especially robust outside two cylinder front end. - St Rollox also impressed by usefulness of Drummond 2-6-0s on the hated G&SW. - Pre-grouping, Caledonian works out outside-cylinder 2-6-0 - Post-grouping, Hughes impressed by this, but it doesn't fit English gauging, so he re-designs from scratch to the same concept but with Horwich practice. - After Hughes retirement, Derby looks to modify, but in the end just adds Midland tender. Presumably Anderson must have ordered the Crabs, otherwise they wouldn't have been built!