An example, consider car tires. In the search for better performance there are numerous ideas (and of course a huge numbers patents) - different compounds, special shapes, apparatus inside the tire, etc etc.
My research team discovers an incredible, really remarkable thing - if you apply microscopic amounts of béchamel sauce to a tire while it is in use, incredibly, the performance of the tire triples in all respects - grip, wear, safety, durability, etc.
(The guys have even worked out how much you need for optimum results with this astonishing discovery - it's 1 pint of béchamel sauce per 10,000 miles of tire use.)
So that's a genuinely amazing discovery or finding - great.
You then have to actually apply the tiny amounts of béchamel to the tires on a car. The team thinks up three nice ideas for that: (i) an ingenious ring of tiny béchamel nipples inside the tire protruding slightly to the tire surface; (ii) an ingenious microscopic pastry brush system which is suspended on the vehicle inside the wheel arch; (iii) an ingenious attachment to the axle hub which rotates with the axle.
So, the basic discovery is scientifically astonishing and took the research team five years and 20 million bucks to stumble on.
The three "devices" to achieve béchamel micro-application are really not a big deal, they just tossed them off, tried a couple prototypes - indeed, its likely you could easily think of other clever application mechanisms.
Q1. Could the patent application in fact just be for the raw "discovery". Hence, the patent application would essentially say "To triple the grip and service lifetime of an automotive tire, we apply x mg of béchamel sauce per km to the tread whilst the tire is operating." You don't even mention the (sundry obvious) ways, that could be achieved.
Q2. If the answer is basically "No, you have to have the device as such" ... then could all three of the device-methods (and perhaps more) be included in the one patent? Or would each of those need its own application? Or?
Q3. If the latter, that seems odd as in some sense you're not really getting any protection on the "actual discovery" - ? (ie, a thousand other patents would immediately be filed with offhand ideas for applying the béchamel sauce - the béchamel sauce being the "actual meaty discovery" at hand - which took years and millions in research to discover.)
What's the deal on this? Is it actually even possible to patent such a "discovery" as in the example?
(I mention USA for formality of the question, but of course issues in any jurisdiction would be of interest.)