The duration was in the operators but not in the experiment: it was experienced, but it disappears when the results of the experiment are conceptualised. A succession of events is in ourselves and not in the events observed. If a point is said to move along the locus _OD_ through the positions _A_, _B_, _C_, it is we that have the feeling of succession, and the whole trajectory, or locus, or path of the point corresponds with a portion of our duration. The operation of boiling the kettle corresponds with a portion of our duration, which in its turn corresponds with that part of our duration which was marked by the positions of the hands of the clock. Thus we perceive a simultaneity in these two trains of events, and this enables us to assign a certain period of astronomical time to the operation of raising the temperature of the water, in the conditions of the experiment, from 17° C. to 100° C. But there is nothing absolute in this interval of astronomical time: what is absolute is that certain successions of events always correspond with other successions of events. A certain number of swings of a seconds-pendulum always corresponds with a certain rise in temperature of a definite mass of water which is in thermal contact with an indefinitely large reservoir of heat at a certain temperature, and, no matter how often we repeat this experience, the same simultaneity is always to be observed. Thus what the physicist considers is not intervals of his own duration but series of correspondences--that is, correspondences of certain standard events with the events which he is studying.