The recovery on a sliding seat is not quite the counterpart of that on a fixed seat. On the fixed seat the recovery should be the converse of the stroke: i.e. the arms, which came in latest, while the body was still swinging back, should shoot out first, while the body is beginning its return swing; and just as the first part of the stroke was performed with straight arms and swinging body, so the last part of the recovery should disclose a similar pose of arms and body. But upon a slide there is not exactly such a transposition on the recovery of the motions which are correct for the stroke. The hands play the same part as before; they cannot well be too lively off the chest and in extension, because the knees require more clearing on slides, and the sooner the hands are on the safe side of them the less chance is there of fouling the water on the return of the blade. But, as regards the relations between slide and swing, these should _not_ bear the same relation conversely which they did to each other during the stroke. The pupil was enjoined not to let his slide run ahead of his swing while rowing the stroke through; but on the recovery he may, and should, let his slide get well ahead, and be completed before the body has attained its full reach forward. The body should not _wait_ for the swing to do its duty first, but it should begin at once to recover, though more leisurely than the legs. The reasons for this are:-- 1. The pace of the slide lends impetus to the trunk, and eases the labour of the forward swing; it transfers some of the exertion of recovering the trunk from the abdominal muscles, which are weak, to the flexors of legs and loins, which are much more powerful, and are better able to stand the strain. 2. The body needs some purchase upon which to depend for its recovery, and the legs can aid it in this respect much more effectually when bent than when rigid. Therefore, since staying power is greatly affected by the amount of exertion involved in recovery (as explained in previous pages), the oarsman will last longer in proportion as he thus omits the recovery of his trunk, by accelerating his slide on the return. Many good oarsmen slide until the knees are quite straight. In the writer's opinion, this is waste of power: the knees should never _quite_ straighten; the recovery is, for anatomical reasons, much stronger if the joint is slightly bent when the reversal of the machinery commences (No. 4, p. 111). The extra half-inch of kick gained by quite straightening the knees hardly compensates for the extra strain of recovery; also leg-work to the last fraction of a second of swing is better preserved by this retention of a slight bend, and an open chest and clean finish are thereby better attained. Engineers, who know what is meant by a 'dead point' in machinery, will at once grasp the reason for not allowing the legs to shoot quite straight.