The common distinction between the low dark maria and high brighter continents, readily made on lunar photographs, must be refined. The bright areas are of two types: (1) accreted crust, modified by a sequence of impacts that led to the formation of the familiar large lunar craters; and (2) maria or accreted crust, buried by material clearly ejected from Mare Imbrium and a few smaller splash centers. The major mountain ranges (Alps, Caucasus, Apennines, Carpathians) belong in this second category; they seem to be composed of light‐colored blocks of crust, mixed with large quantities of darker lava, now solidified, which look somewhat like mud flows.The recent suggestion by T. Gold and others that the maria are featureless because they are Med with a very deep layer of dust is contradicted by several groups of observations: (1) The maria, seen under very low Sun, show an intricate fine structure, including discrete flows, minor faults, graben, etc. (2) Impact craters situated on the ‘shore’ lines of maria are often tipped toward the mare, like partly‐sunk boats lying against the shore line of a lake. Apparently, the foundations of these crater mins were melted locally and the ruins have subsided or even submerged. (3) A compelling argument is derived in the next paragraph from the nature of the ridges on the maria. (4) There is no trace of a filling effect of very small craters such as a thick, mobile dust layer would cause. Actually, the maria are beset with hundred of small craters iabout 1 km in diameter and smaller), whose numbers and shapes indicate that they are meteoritic in origin. Large craters, (10 to 100 km), as found on the continents, are nearly absent on the maria. These large craters appear to represent an early phase of lunar development, preceding the development of the maria and partly concurrent with it. The maria were probably formed about 4.5 billion years ago.