Follow the oil
from the sump.

The factory workshop PDF, p. 10 shows the pump removal and overhaul sequence and the limits above on an Engine—U.K. Specification page. A Midget 1500 replacement listing and replacement pump photograph help explain the visible body and pickup form. They do not identify the pump fitted to your North American car. The CAD pump position, rotor profile, bore, shaft length, bolt centers, tube bend and strainer clearance are estimates. Its shaft is not connected to the distributor drive in the model.

The linked factory scan prints both inner and outer rotor end float as 0.0004 in (0.1 mm). These units disagree: 0.0004 inch converts to 0.01016 mm, while 0.1 mm is approximately 0.00394 inch. The Spitfire 1500 Italian factory manual, RTC9237, PDF p. 62 separately prints a 0.102 mm maximum axial rotor gap, with a straightedge across the pump body. This supports approximately 0.1 mm as the intended scale—an inference, not a Midget correction notice. Its other printed limits are 0.254 mm between rotors and 0.203 mm between outer rotor and body. These remain Spitfire comparisons; confirm the applicable North American Midget specification and fitted pump before using a service limit. The same page checks rotor end float with a straightedge across the pump body face; gaps in the illustrative CAD assembly do not reproduce that measurement.

The factory overhaul drawing places the cover and strainer opposite the projecting drive shaft and removes the cover complete with strainer. The current CAD follows that order: the shaft exits through the housing above, the cover closes the chamber below, and a straight pickup joins the cover to an illustrative strainer neck. The vertical-pickup replacement listing supports this comparison layout; it does not identify the fitted pump.

The native model diagnostic ↗ checks modeled cover-to-body and pickup-to-cover face contact without material overlap. It still extracts a 36.2 mm housing bore and 34.8 mm outer rotor, coincident rotor axes, and 2 mm gaps above and below the disks. Those are dimensions of unmeasured recognition geometry, not operating clearances or end-float readings. Casting columns, pickup threads, locknut, actual strainer construction, pressure-side oil passages and drive engagement remain unresolved. The estimated pump mounting columns, cover ears and hardware overlap the estimated sump wall by approximately 270.6, 103.6 and 51.9 mm³ respectively. These are reconstruction conflicts, not evidence of interference on the car. The fitted mounting station, casting and sump section must be surveyed before an installed-fit or replacement-path claim. The report applies to the fingerprinted native source, not an arbitrary owner-loaded model. Select a pump or sump part in the CAD viewer and choose Compare pump and sump to see their source positions together. The sump becomes translucent and explode is paused; the cutaway controls can expose the estimated wall. The view preserves overlaps rather than moving parts to clear them. Current exports also carry seven per-build sump-casting comparisons: body, cover, both rotor groups, mounting bolts, pickup and strainer. Their individual overlap volumes and model minimum distances appear only after all eight original source meshes match that export’s native fingerprints. The rows are not summed into a total. Older, incomplete or mismatched owner exports remain explicitly unchecked; the public result is never substituted. Pan gasket, fasteners and drain parts are context only, outside these seven pairs. Browser notes and visual cutaways do not recalculate the native source comparisons. Rebuild the diagnostic separately from the original owner measurement log using the native audit instructions; that retains source evidence and checks the reconstructed owner sump, without auditing an arbitrary browser-loaded mesh. Older exports without the current layout record need rebuilding.