In following the history of geography we see the Cassipa, figured as a rectangular parallelogram, enlarge by degrees at the expense of El Dorado. While the latter is sometimes suppressed, no one ventures to touch the former,* which is the Rio Paragua (a tributary stream of the Caroni) enlarged by temporary inundations. (* Sanson, Course of the Amazon, 1680; De L'Isle, Amerique Merid. 1700. D'Anville, first edition of his America, 1748.) When D'Anville learned from the expedition of Solano that the sources of the Orinoco, far from lying to the west, on the back of the Andes of Pasto, came from the east, from the mountains of Parima, he restored in the second edition of his fine map of America (1760) the Laguna Parime, and very arbitrarily made it to communicate with three rivers, the Orinoco, the Rio Branco, and the Essequibo, by the Mazuruni and the Cujuni; assigning to it the latitude from 3 to 4 degrees north, which had till then been given to lake Cassipa. I have now stated, as I announced above, the variable forms which geographical errors have assumed at different periods. I have explained what in the configuration of the soil, the course of the rivers, the names of the tributary streams, and the multiplicity of the portages, may have given rise to the hypothesis of an inland sea in the centre of Guiana. However dry discussions of this nature may appear, they ought not to be regarded as sterile and fruitless. They show travellers what remains to be discovered; and make known the degree of certainty which long-repeated assertions may claim. It is with maps, as with those tables of astronomical positions which are contained in our ephemerides, designed for the use of navigators: the most heterogeneous materials have been employed in their construction during a long space of time; and, without the aid of the history of geography, we could scarcely hope to discover at some future day on what authority every partial statement rests.