Perhaps no other French poem of recent times has been so much discussed, commented on, and analyzed as Arthur Rimbaud's Le Bateau Ivre. Hence it may seem presumptuous on my part to propose to the reader another lengthy discussion of a work about which “everything has been said.” If I do so, it is because of a conviction that much remains still to be said. For whereas there have been proposals for the total reading of the poem which seem to be generally acceptable, these usually fail to do more than suggest the broad general meaning; they do not provide insights into specific passages or solutions to the many difficulties of detail. Moreover, they tend to concentrate on the problem of meaning to the exclusion of the problem of structure, and it is in the problem of structure that I am especially interested. I should wish to discover, in the following pages, not only a more nearly complete reading of Le Bateau Ivre but also the answer (or part of the answer) to a theoretical question that constantly arises in the analysis of symbolist poetry. The problem simply stated is this: How much of a symbolist poem is actually involved in the central symbol about which the poem is built, and what purpose is served by such parts of the poem as are not involved in that symbol? Hence my subtitle, “The Limits of Symbolism,” which does not mean to impose limits upon the poet or to suggest that any such exist in the nature of things, but merely to inquire into the operation of these limits in the case of Le Bateau Ivre.