Clock speed management in a modern CPU is most of all thermal management: Starting from a base frequency, that can be sustained with optimal cooling over long periods of time, the CPU can be "boosted" to higher clock speeds if either not all parts of the CPU are used (e.g. single-threaded workloads) or throttled (if the cooling is unable to remove the heat, most notoriously in Apple "Pro" Books).
Your Screenshot gives the impression of a hexacore CPU with only one core running full steam ahead. In this case the thermal management is able to boost the clock frequency from the base 3.2 GHz to over 4.2 GHz: While the one core produces a lot of heat, the others are basically idle, most likely partly switched off. This clock speed can not be sustained indefinitly, and when another core has to pick up some load (background tasks etc) the boost on the active core has to be reduced.
Since the boost/throttle are reevaluated from thermal and load datapoints many times per second (100s of times IIRC), the current clock speed will fluctate quite heavily. This is normal and to be expected.