The positioning time or random access time, consists of two parts: the time…

Concept. Disk access time is not a single quantity. It is positioning time — also called random-access time — followed by the transfer time in which the data actually moves. Positioning time exists because a movable-head disk must align the head along two independent axes before any byte can be read: radially, across the cylinders, and angularly, around the spinning platter. Each alignment has its own named delay, and positioning time is their sum.

• Radial alignment: the actuator arm travels across the platter until the head sits over the required cylinder. This delay is called seek time.

• Angular alignment: the platter keeps spinning until the required sector passes under the head. This delay is called rotational latency.

Applying it here. The stem quotes the first alignment word for word — "the time necessary to move the disk arm to the desired cylinder". That is the radial component, so the blank is seek time. The half of the definition the stem stops short of is rotational latency, which is why the sentence says positioning time has two parts.

Cross-check. The two remaining terms measure different things entirely:

• Transfer rate is a throughput in bytes per second, and it applies only after positioning has finished — once the head is already over the target sector.

• Flash drives are a storage technology rather than a time interval, so positioning time does not decompose into these two parts for them at all.

So positioning time = seek time + rotational latency. The arm-movement half named in the stem is seek time. This two-part decomposition is the standard definition given in Operating System Concepts (Silberschatz, Galvin and Gagne) in the disk-structure chapter.