It is shown that proton flares and loop-prominence systems form in the same type of active regions. Quite often both these phenomena appear simultaneously, but there are also many cases, when only the proton flare or only the loop-prominence system fully develops. The proton-flare active regions are not randomly distributed on the solar disk, but they tend to occur in complexes of activity which stay on the solar surface for many months and even years. Attention is called to the peculiar clustering of proton-flare regions on the Southern hemisphere, where two sources of activity, at a longitudinal distance of about 180°, seemed to move on the solar disk between 1956 and 1962 opposite to the solar rotation, shifting in the longitude at about 70 heliographic degrees per 10 solar rotations.In my contribution, I would like to discuss very briefly two problems -the oc currence of loop-prominence systems in active regions which produce proton flares, and the occurrence of such proton-flare active regions in complexes of activity.In 1964, Bruzek called attention to the fact that all loop-prominence systems observed on the disk were associated with proton flares, and he concluded that this association of loop-prominence systems with proton flares was a general character istic of these two active phenomena (Bruzek, 1964).We have tried to verify this conclusion of Bruzek using the catalogue of flares associated with type-IV radio bursts prepared by Olmr and myself (1966). This catalogue, containing 174 events, can also be considered for a list of proton flares which appeared on the Sun from 1956 to 1963, and we compared it with 65 loopprominence system occurrences, taken from lists prepared by Bruzek (1964) and Kleczek (1967). We have verified that all 24 loop-prominence systems observed on the disk were preceded by proton flares listed in the catalogue, in full agreement with Bruzek's results. Of course, one must not forget that Bruzek tried to find loop promi nences in this type of flares and, therefore, he might perhaps have missed some other events. For the limb-prominence systems the situation is quite different. Only 9 events of the 40 observed limb systems were clearly preceded by proton flares. It is true that, due to the directional sensitivity at long wavelengths, the classification of type-IV bursts for flares close to the limb is difficult, but one can hardly believe that we could have missed 31 type-IV bursts out of the total number of 40. And this is also sup ported by the fact that 14 events of these loop-prominence systems were observed on the Western solar limb without any PCA effect, which accompanies the proton flares