The driver stages get their input from the 13560 kHz oscillator.
The output voltage of the oscillator is approx. 400 mV peak-to-peak
(140 mV eff) if left open. The source resistance of the oscillator is
approx. 300 Ohms. The signal at the input of the driver stage 1 is 
approx. 210 mV peak-to-peak (75 mV eff) if connected. At the output
of the driver stages 1+2 (emitter of second transistor) you may
measure approx. 2.4 Volts peak to peak (=0.85 V eff) if left open.
With a load resistor of more than 100 Ohms, the source resistance
of the driver stages 1+2 at its output is about 15 Ohms. With a load
resistor of 150 Ohm at the emitter of the second transistor you get
approx. 2.2 Volts peak-to-peak (0.75 V eff). Maximum power output
of the driver stages 1+2 is less than 5.5 Milliwatts/100 Ohm or
4 Milliwatts/150 Ohm. At the emitter of the second transistor, the
load resistor should not be below approx. 100...150 Ohms because 
otherwise distortion appears and increasing harmonics are produced.
Without a load resistor at this point of the circuit, the second 
harmonic (27120 kHz) is about 60 dB weaker than the fundamental 
(13560 kHz), according to approximation using the s-meter of my 
Grundig satellite 3400 prof. radio.
   The following stage (with transistor BC549B) is the AM modulator.
The collector-emitter conductance of the transistor varies with the
course of the modulating wave form which is input to the base of the
BC549B, thus modulating the amplitude of the RF signal. The collector-
emitter conductance leads more or less parts of the RF signal to
ground. Limitations of AM modulation with this modulator are (1)
the limited maximum collector-emitter conductance as
represented by U(CE sat) of the BC549B which is 0.2 Volts and not zero
(preventing the U (RF) from becoming zero and thus limiting the
grade of modulation to approx. 80%), (2) the linearity of the
modulation which is not perfect (bringing little distortion to the
modulation and thus reducing tone quality a bit), (3) the asymmetric
conductivity as to the positive and negative part of the RF wave
(producing distortion of the RF waveform which leads to an increase
in second order harmonics), and (4) the reduction of Rf signal voltage,
making a further stage neccessary.
   The last stage is the transistor 2N2219A. Several measures and steps
have been taken to reduce harmonics. At the output, the second harmonic
(27120 kHz) is about 40...60 dB weaker than the fundamental. Maximum 
output peak to peak voltage without load is about 2.2 V (0.75 V eff),
if the carrier control is put to maximum. For proper AM modulation, 
carrier control should be adjusted to an output voltage which is half 
the maximum. Source resistance at the output is about 200 Ohms. With
carrier control adjusted to an output voltage of 1.1 V peak to peak
(370 mV eff), and using a load resistance of 200 Ohm, the output power
is less than 1 Milliwatt. Output power depends on the LC-network you
will use to connect your antenna. To my opinion, whatever you may 
connect, the maximum output power of this circuit could never exceed
10 Milliwatts.
   Connection of antenna. Due to legal limitations, I'm aware I cannot
tell you of any significant experience. As far as I could try these
things without getting in conflict with german rules, I found that
a very short antenna (less than 1 meter) is best tuned by just a coil
of 10 µH between output and antenna, whereas the capacitance between
antenna foot and ground should be not more than 0...2 pF. The longer 
the antenna wire, the lower the coil impedance should be, and the 
higher the capacitance should be which you put between antenna foot 
and ground. For example, with 3...4 meter, the best impedance was 
between 3.3µH and 6.8 µH. I recommend that with your particular long 
antenna that you may have, first try different coils round about 
1 ... 2.3 µH and leave out any capacitor. If possible, use a coil with
an adjustable ferrite rod inside. After you found the coil with the 
highest field strength, then put a variable capacitor between antenna 
connector and ground and try to find a good compromise between high 
field strength and low second harmonics.