The relation of speaking and listening in second language acquisition has not been clarified. Some writers maintain that the child should first listen to a foreign language so as to be able to distinguish the sounds before beginning to speak utterances in the new language (e.g. Brooks, 1960). On the other hand, Lane (1964) has presented evidence to suggest that the child should begin by speaking the language in order to facilitate more careful discrimination among the sounds. Keislar and Mace (1965) have indicated that, since many different definitions of listening and speaking are used, the problem should be defined in terms of tasks which clearly specify the instructions, the stimulus and the response required. Mace (1966) conducted a study with primary children in which he attempted to throw some light on certain aspects of this problem. In his investigation he had four groups who were given fifteen minutes of French instruction each day for two weeks. One group was taught, during the first week, to speak French utterances appropriate to projected pictures and then, during the second week, to listen to and respond with understanding to the utterances they heard. A second group was given the same instruction in the reverse sequence: a week of training in listening comprehension followed by a week of speaking training. A third and fourth group received concurrent training combining both listening and speaking in fifteen-minute daily sessions over the two-week period. All four groups were given the same 600 frames of French instruction; the differences among the treatments lay entirely in the order in which the blocks of speaking and listening were presented.