Most of the dormancy in rice seed can be accounted for by the inhibitory influence of the husk, and most of the residual dormancy after dehusking can be attributed to the inhibitory influence of other covering structures—either the pericarp or testa, or both. It is shown that the rate of water absorption is the same in dormant and non-dormant seeds and that dormant seeds are capable of absorbing sufficient water for germination. The covering structures therefore do not cause dormancy by restricting the entry of water. Removal of a small area of the husk breaks the dormancy of a large proportion of the seeds; but for some seeds this treatment is ineffective whereas removal of the entire husk would break dormancy. The site of the excision of a small area of the husk can alter the effectiveness of the treatment: removal of a portion of husk immediately over the embryo is no more effective than excising a similar portion nearby, but the removal of part of the husk some distance from the embryo is not as effective. Sealing the perforations with paraffin wax has little effect except when carried out as soon as possible after the excision is made, and then only in positions distant from the embryo. Attempts to extract a water-soluble or ether-soluble germination inhibitor from the husk and other parts of dormant seed or to demonstrate the presence of inhibitors by indirect methods have not been successful. Nor has it been found possible to extract a water-soluble germination stimulator from seed which has broken dormancy. The implications of these results are discussed.