The treatment of time is more thoughtful. Time is inseparably connected with movement or change. We only perceive that time has elapsed when we perceive that change has occurred. But time is not the same as change. For change is of different and incommensurate kinds, change of place, change of colour, &c.; but to take up time is common to all these forms of process. And time is not the same as motion. For there are different rates of speed, but the very fact that we can compare these different velocities implies that there are not different velocities of _time_. Time then is that in terms of which we _measure_ motion, "the number of motion in respect of before and after," _i.e._ it is that by which we estimate the _duration_ of processes. Thus _e.g._ when we speak of _two_ minutes, _two_ days, _two_ months as required for a certain process to be completed, we are counting something. This something is time. It does not seem to occur to Aristotle that this definition implies that there are indivisible bits of time, though he quite correctly states the incompatible proposition that time is "made up of successive _nows_," _i.e._ moments which have no duration at all, and can no more be counted than the points on a straight line. He recognises of course that the "continuity" of motion implies that of time as well as of space. Since, however, "continuity" in his language means the same thing as indefinite divisibility, it ought not to be possible for him to regard time as "made up of _nows_"; time, like linear extension, ought for him to be a "length of" something.