In this article, a historical review of discussions and controversies about the characteristics and the definition of "graywacke" is illustrated. Definitions of graywacke have proliferated owing to divergence of aims and to a somewhat unavoidable logical confusion in the arguments. The most important of the "graywacke problems" is the "matrix question". From this point of view, HUCKENHOLZ's description (1963) that the original graywacke in Harz Mountains is poor in matrix is very noteworthy. DOTT (1964) made a very reasonable discussion on the graywacke problem, especially on the classification of immature sandstones. The question about the origin of matrix, however, remains unsettled. CUMMINS (1962) has emphasized that there is a direct correlation between the percentage of matrix and the age of graywacke, and argued that the most of the matrix material in the graywacke result from diagenetic breakdown of unstable sand grains. His paper is important as the first discussion based on a geohistorical standpoiut. Lately, however, HOLLISTER and HEEZEN (1964) found very muddy sands (recent graywacke-type sands) from deep-sea core-samples. These data offer a serious objection to CUMMIN's opinion. Diagenetic origin of some matrix is undeniable ; but the diagenefic and authigenetic alteration of minerals causes the grain size of matrix matter to increase rather than to decrease. Scarcity of apparent matrix of Harz graywacke may be explained from this reason. The greater part of matrix of many graywackes is presumably detrital in origin.