I thought that HPX42 was a Philips model number, so I searched for "HPX42 Philips" without quotes. I didn't notice anything relevant, but I must have missed a link. Here is a *very* helpful link:
http://onninen.procus.fi/documents/original/13061/7/0/5021.pdf. Googling your search term without quotes yields a table with nearly the same kind of info in this very forum:
http://www.candlepowerforums.com/vb...bulb-ratings&p=3814026&viewfull=1#post3814026. Both show HPX42 as 4.7V, base P13.5s. One has to be careful, as local hardware stores have identical looking bulbs that require approximately half the voltage (Krypton, I think). Having searched around locally, I think it might be wise to just search Amazon, as I probably burn up many times the cost of gas just to find a suitable bricks-and-mortar source.
As a fallback plan, the last place I visited suggested a local establishment, "Buchanan Lighting".
Based on the opinions of staff at the 1st last place I visited, I might be better off finding an LED with the same base and voltage. They are brighter and the batteries will last longer. Being an electrical engineer by training however, I was initially concerned that the right voltages might be hard to find. My flashlight serializes 4 AAA batteries, yielding 6V unloaded, and 4.8V loaded (the latter from an earlier respondent on this thread). I would need a LED bulb that accepts this kind of voltage source and loads it in a similar manner. I'm naive about batteries and flashlight bulbs, but if a bulb is designed for a 4-cell, then its loading effect should take into account the output resistance of the batteries, so the only important parameter is number of cells it was designed for. The exact loaded voltage will depend on the bulb design, and whether that yields the right operating voltage is the bulb designer's problem.
The take-away is that I only have to match the cell-count and the base specification. I then have the freedom to choose things like LED rather than xenon or krypton. I took the above candlepowerforums table and dumped it into a spreadsheet to filter by cell-count. I found that the options are all gas-filled. Again, I'm new to the standards in flashlight bulbs, but I assume that these are all incandescent. So I have some searching ahead of me. Before doing so, I'll check the bulb with a multimeter, as suggested, to make sure that it is indeed the bulb rather than the flashlight. Otherwise, I'm just wasting my time (though it has been educational).
By the way, can anyone comment on how normal it is for LED bulbs to be designed with compatible bases and unloaded voltages as incandescent bulbs?
AFTERNOTE: Darn, ohmeter test shows that the filament in the bulb is not broken. It's essentially a short, which jives with my understanding of incandescent bulbs. The filament only increases in resistance when it heats up. So this means that the problem is elsewhere. A voltmeter test of the 4-cell batter assembly shows 5.88 V, which is right. So the problem is the pushbutton switch or the conductors connecting the battery assembly to the bulb. If I can't find the problem and fix it, I will graduate to an LED flashlight. Thank you all for educating me thus far.
FINAL SCOOP: It was difficult poking the ohmeter probes down there, but the problem turns out to be a bad connection between the spring at the bottom and the metal strip that goes down there from the push-button switch. From visual inspection, the joint looks terrible, speaking from someone who has breathed in lots of solder flux in another life. Really badly made. An environmental crime. The makers are off-shore, but I'm thinking that the manufacturing *process* was not designed for robust construction.
Since I'm going to graduate to an LED flashlight, there isn't even any point in saving the bulb.