While holding that the currents of the ocean form one system of circulation, we must not be supposed to mean that the various currents are connected end to end, having the same water flowing through them all in succession like that in a heating apparatus. All that is maintained is simply this, that the currents are so mutually related that any great change in one would modify the conditions of all the others. For example, a great increase or decrease in the easterly flow of antarctic water in the Southern Ocean would decrease or increase, as the case might be, the strength of the West Australian current; and this change would modify the equatorial current of the Indian Ocean, a modification which in like manner would affect the Agulhas current and the Southern Atlantic current—this last leading in turn to a modification of the equatorial current of the Atlantic, and consequently of the Brazilian current and the Gulf-stream. Furthermore, since a current impelled by the winds, as Mr. Laughton in his excellent paper on Ocean-currents justly remarks, tends to leave a vacancy behind, it follows that a decrease or increase in the Gulf-stream would affect the equatorial current, the Agulhas current, and all the other currents back to the antarctic currents. Again, a large modification in the great antarctic drift-current would in like manner affect all the currents of the Pacific. On the other hand, any great change in the currents of the Pacific would ultimately affect the currents of the Atlantic and Indian Oceans, through its influence on the Cape Horn current, the South Australian current, and the current passing through the Asiatic archipelago; and _vice versâ_, any changes in the currents of the Atlantic or Indian Oceans would modify the currents of the Pacific.