T1 is just a VHF Colpitts oscillator.
Any components up to the coupling capacitor C9 appears to be a weird setup in my opinion because a standard FM transmitter setup will have the input at the NPN's base and the output at the collector.
The varicap diode's capacitance plus C9 in series with it determines the RF frequency. Since the varicap diode's capacitance changes with voltage changes, then VC1 tunes the RF frequency and the input audio modulates the varicap diode's voltage providing pure FM.
L1 and L2 make transformer coupling. I don't use transformer coupling personally.
L2 has fewer turns than L1 so the transformer matches the high impedance at the collector of T1 to the low impedance of the base of T2.
I think VC1 and VC2, L3, L4 and C7 make one or more filters to prevent interference from happening at the wrong stations.
VC1 adjusts the RF frequency of the oscillator by changing the capacitance of the varicap diode.
VC2 adjusts the RF power output level. Lower resistance equals higher power.
L3, VC1 and VC2 match the impedance of the collector of T2 to the antenna.
L4 is just an RF choke with C7 a bypass cap to keep high power RF away from the supply for T1.
The 100K resistor limits the signal, but I think 100K is too high.
R2 doesm't limit or reduce the audio signal since the varicap diode is reverse-biased and therefore is a very high impedance, and the tuned circuit at the collector of T1 is also a very high impedance.
If R2 was less than 100k then it would reduce the sharp tuning.
Try it. It allows an FM transmitter to make pure FM without the AM added by simpler circuits. :lol: