Further proof -- as if it were needed -- that I have way too much time on my hands...
The third photo shows the very small pinhole on top of the flat red upper part of the reservoir. In my opinion to let in air that is mixed with the water. Height from this part of the top to the fountain spray is 55 mm.
Many thanks for the original directions, photos, and dimensions, Lourens. Per the directions, the "pin hole" is a lubrication port leading to the bearing supporting the rotor.
Oil port: I'm betting that this oil port has no significant effect on the air and water flows through the top at any stage. In the cut-away drawing, the "rotor bearing" surrounds the centerline above and between labels "C" and "R". When lubricated, the metal bearing surfaces would be separated by a thin film of pressurized oil. Given the weight of the rotor, this film would be unlikely to pass either air or water. Ditto for any water film or metal-on-metal contact in the absence of oil.
So, if there are no other holes above the vertical water intake at the bottom of the rotor (above label "B"), that intake must double as the only air intake of significance. Let's just call it the "intake" for short.
Spin-up after immersion? The directions clearly instruct the user to spin the top
before immersing the intake, as Cyril did in his video. Whether this order is a necessity or just a matter of convenience is unclear. If a necessity, then the air fountain stage would be an essential step toward the water fountain stage, with a brief air-powered priming stage in between.
The test would be simple enough: Submerge the intake first and
then spin up the top.
The drawing cross section shows a tube wall going vertical some distance before reaching the reservoir. The French description says that vacuum is created inside that suctions the water.
Agree, close inspection of the cut-away shows that this lowermost vertical rotor section surrounds an annular intake already filled with water. The water forms a continuous peripheral layer from there to the fountain inlet at label "C". To my mind, this scenario depicts only during the fully primed water fountain stage.
The inner wall of the intake is a cylindrical "cup" holding the (stiff ball-and-socket?) joint that allows the top to pivot on the foot. The cup is rigidly attached to an internal "spindle" ending above in the rotor bearing. The radial fountain inlet arm "C" is clearly part of the spindle, which is also rigidly attached to (i) the vertical fountain outlet tube ending above near label "D", and (ii) the handle via (i). As ta0 suggested, the rotor spins relative to the spindle and all its extensions. How the manufacturer got that L-shaped spindle part inside the ostensibly one-piece rotor is beyond me.
Self-priming? The old book explanation ta0 posted clearly implicates air in the priming process, but the author could have been mistaken. If the external water bath were deep enough to submerge the entire vertical intake and a portion of the sloping rotor reservoir as well, the fountain could conceivably be self-priming in the manner Dick suggested, with no need for a priming air flow.
However, the water in Cyril's video doesn't look deep enough for that to me. And judging from Lourens' measurements and the cut-away, neither is the design water bath depth of 1/2 inch (12.7 mm) called for in the original directions. If so, the role of the priming stage would be to propel water through the vertical intake and into the sloping rotor reservoir, where centrifugal force can take over. The most likely propellant in this case would be the air flow already established during the air fountain stage.
External Archimedes screw? The odd spiral ridge on the outer surface of the vertical intake could be strictly ornamental, but it could also assist in priming by acting as an Archimedes screw. The external water lifted by the screw would increase local water pressure at the submerged intake, thus reducing the net head to be overcome during priming.