This is a long post putting together several threads of evidence that the Rossi reactor is a cavitation reactor.
Pressurized hydrogen in contact with nickel nanopowder is cavitated, eventually undergoing first chemical and then nuclear reactions, with copious excess heat. The Casimir effect could be involved in a novel way for the chemical part. Fractional hydrogen could be involved. The nuclear part is confused by conflicting reports about radioactivity. Best bet is that it is the "Ra Reaction" of our sun - conversion of hydrogen into deuterium, but that is speculation for other posts. An appreciation for the history of the Griggs pump is necessary to grasp all of this. Here is some information on the pump, which is still in commercial production in Rome (the one in Georgia): http://www.rexresearch.com/griggs/griggs.htm The Hydro-Dynamics pump employed cavitation and shock waves from a dimpled rotor spinning inside a housing to increase the temperature of water flowing through the device. It was tested on a number of occasions to be OU, but not reliably. Jed Rothwell has reported on it, as did Infinite Energy. Now - imagine the rotor being non-rotating ! Cavitation in the Rossi device could be described as Griggs pump - with the reactor substituted for the dimpled rotor. The reactor cavitates violently, but at low excursion, and would not be noticed in a demo, since the effects are dampened by the water flow. Primarily, it produces cavitation INSIDE the cell, and ironically this would never have been noticed outside the cell except for contrasting the two tests in Bologna, one with low water flow, and one with high. This could be a most fortuitous discovery for anyone working on a replication. The higher water flow is substantially more efficient than low flow - and the reason for that relates to Griggs and to optimizing cavitation. The outer surface of the reactor would be the functional equivalent of a transducer to cavitate the water flowing over it, but only if the water pressure was high enough. The Griggs pump needs massive water flow. The internal cavitation is not changed much in either case, so there is always that base level of heating due to internal cavitation. The "clue" that cracks the case - turns out to be in contrasting the two test results in the context of a heated argument. Magnetostriction is a property of ferromagnetic materials, particularly nickel - that causes them to change shape during magnetization. It was discover by Joule himself, in nickel, long ago. The effect is responsible for the familiar "hum" which can be heard near transformers at 100 (=2*50) or 120 (=2*60) hertz, plus higher harmonics. Transformers are iron based, but nickel nanopowder could be much more extreme in the effect. With the very high excursion of surfaces at the nano level, the magnetostriction effect would be magnified by perhaps an order of magnitude and yet nearly imperceptible at the local level. The reactor containing the nanopowder would function like a humming transformer core and it could also operate internally with shock waves pushing hydrogen into Casimir cavities. As in the Griggs pump, cavitation generates shock waves which convert mechanical energy into acceleration and eventually into heat energy - in a way that is gainful at times. The Rossi reactor is apparently gainful all of the time, and that could be due to the employment of nano geometry. Many of the common transducers used for sonochemistry are magnetostrictive instead of piezoelectric, as these are more robust at high input. The efficiency is very high. It is too much of a coincidence that the reactor loses it heating effect at a temperature which coincides with the Curie point of nickel, and is more robust when more heat is removed by higher water flow; not to mention that the "resistors" have a magnetic field. An interesting point is that the inventor may have discovered this inadvertently and never thought to optimize the input power, which should be easier to do via an inductive coil instead of resistance heaters. If this proves accurate in a replication effort, then my thanks go out in a general way, to all of the Vortex posters who have mentioned these details in past postings, as I have not had the time to go back to credit them individually - you know who you are :-) Jones