Over the winter I sketched out the idea of a small live steam G3 loco specifically designed for R/C shunting rather than high speed express running; and also to use as many off the shelf components as possible in its construction without compromising overall accuracy too much. The prototype selected was one I had wanted to build for a long time, one of a batch of three 0-6-T locos supplied byDubs to the SM&AR in 1881 which satisfied the requirement of being small and hopefully fairly straightforward. Wheels were available either as castings from Mark Wood or as complete wheelsets from Slaters, and a standard Roundhouse Engineering boiler plus gas tank, burner, regulator and lubricator would all fit the scale outline. Off the shelf R/C gear would hopefully fit in, especially if the side tanks are used. A Paul Forsyth G1 cylinder design (for larger prototypes) would fit well and castings were available although I may well chew them out of stock material. Shunting will require fully reversible valve gear rather than the simpler slip eccentric, and although Roundhouse do supply a simplified Walschaerts set, the original used inside Stephenson's. There is a published G1 design by Martin Evans for a Brighton Atlantic which incorporated Stephenson's, and a quick look suggests it should fit, just. The design seems well proven so presumably the geometry is ok. Finally, and with the shunting brief in mind, there is now an off the shelf 'Slomo' flywheel-type device available which can flatter the slow speed characteristics of 16mm locos for which it was designed, and with the help of the Australian manufacturer a slim version may be available to fit within G3 frames. No official drawings exist for these locos, so far as I know, but Mike Barnsley's excellent book on M&SWJR (which grew from the SM&AR) stock has some elevations so with those plus a selection of photos a working drawing was produced. From that, mainframes were cut, actually milled, from mild steel and buffer beams from brass which happened to be at hand. The distinctive diamond- shaped coupling pocket plates were sweated on, as were rivets into drilled holes. Springing may be proven necessary, or even just desirable, but for the moment simple unsprung axleboxes suffice. Photos just about show progress to date although the frames have since been disassembled again, mounted on the mill at an angle and registers (holes) for the cylinders have been machined in.