A capacitor lets through higher frequencies more than lower ones - the lowest of low, 0Hz is DC ("Direct Current" vs AC-"alternating Current") and that is totally blocked. Think of it as a resistor which value is infinite at DC, and decreases with signal frequency, until it's low enough to be negligible. That "frequency-dependant resistance" is called "impedance".
Smaller values cap need higher frequencies to get to a certain level of impedance. Looking at it another way, a *smaller* value cap will have *higher* impedance for a given frequency, and therefor let through *less* signal.
This is how a treble boost works: have a cap in the signal path (in series) that's small enough to significantly increase in impedance at audible bass levels - it "blocks" bass, and only treble gets boosted. Smaller cap: even less bass. Even smaller: less mids too. Etc. Bigger cap: more bass, up to and including the point your booster is full-range. The cap's value is determined by math (Ohms law rules - literally) and depends on the surrounding circuitry - that's theory for later, now let's tinker.
Now, what happens if you use a BIG cap in series (to let in *all* bass - full range), and stick in a smaller cap to ground?
Remember, that smaller cap will want to let through higher frequencies. THOSE will be shunted to ground. Hey presto, simple bass boost.
Without math, I'd just try a working 1-tranny booster (full range version) and then perhaps stick a 100n from the transistors collector to ground, see what that does, and experiment from there. remember, larger values shunt lower frequencies to ground, higher values shunt higher freq's.
There's no rule you can't try it at the other end of the transistor either - you'll need a different value probably since circuitry conditions are different (there's that pesky theory stuff again).
Build it on a breadboard and experiment.