The problem of turbulent now continues to be an outstanding one in technology and in physics. Of the nine Dryden research lectures so far, four have been on some aspect of the turbulence problem. At meetings such as this one the turbulence problem is always the subject of some sessions and lurks in the background of many others; for example, separated now, combustion, jet noise, chemical lasers, atmospheric problems, etc. It is continually the subject of conferences, workshops and reviews. In his time Hugh Dryden wrote several reviews of turbulent now. In reading some of them again, one statement particularly relevant to the present lecture caught my attention: "-it is necessary to separate the random processes from the nonrandom processes. It is not yet fully clear what the random elements are in turbulent now." Neither is it fully clear what the nonrandom, orderly elements are, but some of them are beginning to be recognized and described. Generally the picture one has had of turbulence is of chaos and disorder, implicit in the name. Although it was known that organized motion could exist, superimposed on the background of "turbulence," for example, vortex shedding from a circular cylinder up to Reynolds numbers of 10^7, such examples were regarded as special cases closely tied to their particular geometric origins and not characteristic of "well-developed" turbulence. It was known that large structures are important in the development of turbulent shear flows and that these ought to possess some definable features. But even when the concept of a characteristic "big eddy" was explored, it was usually in the context of a statistical quantity. The earliest and most decisive attempts to define the form of such large eddies were made by Townsend and his students. In recent years it has become increasingly evident that turbulent shear flows do contain structures or eddies whose description is more deterministic than had been thought, possessing identifiable characteristics, existing for significant lifetimes, and producing recognizable and important events. More accurate descriptions of their properties, how they fit into the complete description of a turbulent flow, to what extent are they central to its development, and how they can be reconciled with the apparent chaos and disorder, are problems which are becoming of interest to an increasing number of researchers. It is the purpose of this lecture to describe some of these new developments. The discussion will draw largely on experiences from our own laboratory; it is not intended to be a complete survey. Other discussions of these ideas can be found in various recent publications.