A critical examination of Freycinet's charts is alone sufficient to shatter the opinion that he utilised the drawings of the English navigator. Had he even seen them, his own work would have been more accurate than it was, and his large chart of New Holland would have been more complete. It has already been shown that the French chart of the so-called Terre Napoleon coasts was in large measure defective, many capes, islands, and bays being represented that have no existence in fact, and a large portion of the outline being crudely and erroneously drawn. Not only so, but if Freycinet had had copies of Flinders' charts before him, use would certainly have been made of them to give greater completeness to the eastern and north-western shores. Flinders, in his last voyage in the Investigator, had made important discoveries on the Queensland coast and in the Gulf of Carpentaria. He had discovered, for instance, Port Bowen and Port Curtis, which had been missed by Cook, had given greater definiteness to the islands near the southern end of the Great Barrier Reef, and had made a dangerous acquaintance with the Reef itself, discovering one narrow alley through it which is marked on modern maps as Flinders' Passage. In the Gulf of Carpentaria he had also done some entirely original work. He had shown, for example, that Cape Van Diemen, represented as a projection from the mainland on all previous maps, was really part of an island, which he named Mornington Isle. Freycinet's charts reveal not the faintest trace of the fresh discoveries which Flinders had achieved around east and north-east Australia, nor do they in any particular indicate that their manifold serious imperfections had been corrected by reference to Flinders' superb charts. In short, the French work, though beautifully engraved and printed, was, in a geographical sense, for the most part too poor to justify the suspicion that Freycinet received aid from the drawings of the persevering captain of the Investigator.