This circuit was used to stop all the BFO drift. The circuit is extremely stable. Turn the receiver off, and then on at any time and temperature, the BFO frequency is exactly the same. The resonator is a Murata CSB 455E. Murata's series number is CSBLA_E 430-509kHz. The data sheet is a .pdf file and the information on this resonator starts on page 28, `Ceramic Resonators(CERALOCK) Lead Type Two-Terminal CSBLA Series`. The last time I checked I didn't find it on the site, but that doesn't mean you can't find it. T1 is a 455kHz IF transformer and may be eliminated if one is not available. This was a circuit I had lying around and had the transformer already wired in the output. It does help to eliminate harmonics from the oscillator.
The FET can be any part number. Some common ones other than the MPF102 are the 2N5486 or 2N4416. The application above covers 452.6 to 455.0 kHz. This is the perfect range for the receiver. The circuit above could be improved to expand the range, but an extended range was not necessary. The amount of drive delivered to the product detector should be on the low side. Driving the product detector too hard will cause leakage to the IF strip and cause oscillations. If you are listening to a CW signal that suddenly jumps around, oscillations in the IF strip from too much BFO drive should be investigated. No problems were experienced with the Resonator/amplifier shown above, but the Hartley (below) with a MOSFET amplifier had to be tamed down. This circuit was inspired by the article "Ceramic Resonators for Cheap and Cheerful VFOs", by Jack Ponton, GM0RWU. Also check out "Mixer VFOs for the HF bands." This article was quite an eye-opener that showed a huge range could be covered by resonators. He shows a circuit that covered 13.990 to 14.280 MHz. The article is well worth reading.