I will only add one more observation to this section. Geologists finding in the most remote period with which we are acquainted, namely in the Silurian period, that the shells and other marine productions{398} in North and South America, in Europe, Southern Africa, and Western Asia, are much more similar than they now are at these distant points, appear to have imagined that in these ancient times the laws of geographical distribution were quite different than what they now are: but we have only to suppose that great continents were extended east and west, and thus did not divide the inhabitants of the temperate and tropical seas, as the continents now do; and it would then become probable that the inhabitants of the seas would be much more similar than they now are. In the immense space of ocean extending from the east coast of Africa to the eastern islands of the Pacific, which space is connected either by lines of tropical coast or by islands not very distant from each other, we know (Cuming) that many shells, perhaps even as many as 200, are common to the Zanzibar coast, the Philippines, and the eastern islands of the Low or Dangerous Archipelago in the Pacific. This space equals that from the Arctic to the Antarctic pole! Pass over the space of quite open ocean, from the Dangerous Archipelago to the west coast of S. America, and every shell is different: pass over the narrow space of S. America, to its eastern shores, and again every shell is different! Many fish, I may add, are also common to the Pacific and Indian Oceans. {398} <Note by the author.> D'Orbigny shows that this is not so. _Summary on the distribution of living and extinct organic beings._