HQM is an optical broad-band photometric monitoring program carried out since Sept.~1988. Our main intention is to search for indications of microlensing in a sample of $\sim$\,100 selected quasars; however, we also want to study the intrinsic variability. We use a CCD camera equipped to the MPIA 1.2$\,$m telescope. Fully automatic photometric reduction relative to stars in the frame is done within a few minutes after each exposure, thus interesting brightness changes can be followed in detail. The typical photometric error is 1--2\,\% for a 17.5\,mag quasar, making HQM the most accurate long-term quasar monitoring program yet carried out. The main results of HQM which we discuss here are: (1) Concerning variability, quasars form two, clearly distinct classes, optically violent variables (OVVs) and non-OVVs. (2) All OVVs are radio loud and probably belong to the blazar class. (3) Non-OVVs have lightcurve gradients of at most several 0.1\,mag\,yr$^{-1}$ in the quasar restframe and can be well fitted by polynomials of low order. (4) Although our data cover only a relatively short timespan, we conclude that there is a large fraction of quasars which would be undetectable in photographic surveys using optical variability alone as the selection criterium. (5) A broad class of flat-spectrum radio quasars are no blazars, they are even less variable in the optical than radio quiet objects. (6) There is some statistical evidence for microlensing in our sample; if present it does, however, lead only to low lightcurve gradients.

HQM is an optical broad-band photometric monitoring program carried out since Sept.~1988. We use a CCD camera equipped to the MPIA 1.2$\,$m telescope. Fully automatic photometric reduction relative to stars in the frames is done within a few minutes after each exposure, thus interesting brightness changes can be followed in detail. The typical photometric error is 1--2\,\% for a 17.5\,mag quasar. We here present lightcurves already evaluated but not shown in Paper I. We also discuss existing literature data.