"To explore the north-west and south-west coasts of Japan," says Kruzenstern, "to ascertain the situation of the Straits of Sangar, the width of which in the best charts--Arrowsmith's 'South Sea Pilot' for instance, and the atlas subjoined to La Perouse's Voyage--is laid down as more than a hundred miles, while the Japanese merely estimated it to be a Dutch mile; to examine the west coast of Yezo; to find out the island of Karafuto, which in some new charts, compiled after a Japanese one, is placed between Yezo and Sachalin, and the existence of which appeared to me very probable; to explore this new strait and take an accurate plan of the island of Sachalin, from Cape Crillon to the north-west coast, from whence, if a good harbour were to be found there, I could send out my long boat to examine the supposed passage which divides Tartary from Sachalin; and, finally, to attempt a return through a new passage between the Kuriles, north of the Canal de la Boussole; all this came into my plan, and I have had the good fortune to execute part of it." Kruzenstern was destined almost entirely to carry out this detailed plan, only the survey of the western coast of Japan and of the Strait of Sangar, with that of the channel closing the Farakai Strait, could not be accomplished by the Russian navigator, who had, sorely against his will, to leave the completion of this important task to his successors. Kruzenstern now entered the Corea Channel, and determined the longitude of Tsusima, obtaining a difference of thirty-six minutes from the position assigned to that island by La Perouse. This difference was subsequently confirmed by Dagelet, who can be fully relied upon. The Russian explorer noticed, as La Perouse had done before him, that the deviation of the magnetic needle is but little noticeable in these latitudes. The position of Sangar Strait, between Yezo and Niphon, being very uncertain, Kruzenstern resolved to determine it. The mouth, situated between Cape Sangar (N. lat. 41 degrees 16 minutes 30 seconds and W. long. 219 degrees 46 minutes) and Cape Nadiejeda (N. lat. 41 degrees 25 minutes 10 seconds, W. long. 219 degrees 50 minutes 30 seconds), is only nine miles wide; whereas La Perouse, who had relied, not upon personal observation but upon the map of the Dutchman Vries, speaks of it as ten miles across. Kruzenstern's was therefore an important rectification. Kruzenstern did not actually enter this strait. He was anxious to verify the existence of a certain island, Karafonto, Tchoka, or Chicha by name, set down as between Yezo and Saghalien in a map which appeared at St. Petersburg in 1802, and was based on one brought to Russia by the Japanese Koday. He then surveyed a small portion of the coast of Yezo, naming the chief irregularities, and cast anchor near the southernmost promontory of the island, at the entrance to the Straits of La Perouse. Here he learnt from the Japanese that Saghalien and Karafonto were one and the same island.