ings which designate the slope and undulation of the ground, afford at the same time a shade to the maps overcharged with names. These names become not unfrequently illegible, when an engraver attempts to produce a grand effect by the distribution of chiaro scuro. Hence the geographer who has carefully discussed the astronomical position of the places becomes uncertain whether he ought to preserve distinctness of character, or render more perceptible the relative height of mountains. One of the most beautiful maps which was ever published in France*, the one drawn up in the war depôt in 1804, sufficiently proves how difficult it is to reconcile two opposite interests, the interest of the geologist and that of the astronomer. The fear of giving too great an extent to my work, and the difficulties attending the publication of an atlas of which no government defrays the expense, made me abandon a project which I had once formed of joining to each section of territory a physical map in a horizontal projection.

* We have discussed in the eighth chapter the extraordinary regularity in the position of the Mexican volcanoes. I am uncertain as to the longitude of the Pic de Tancitaro, which has been twice surveyed from a distance. I fear some error has crept in at copying the angles; but the latitude of this Pic is sure enough to within about eight minutes.

II. MAP OF NEW SPAIN AND THE CONTERMINOUS COUNTRIES ON THE NORTH AND EAST.

I have already explained the motives by which I was induced to curtail my large map of New Spain within too narrow limits for representing, on the same plate, the whole extent of the kingdom from New California to the intendancy of Merida. The second map is destined to remedy this inconvenience. It shows at once, not only all the provinces which depend on the viceroy of Mexico and the two commandants of the provincias internas, but also the island of Cuba, whose capital may be considered as the military port of New Spain, Louisiana, and the Atlantic part of the United States. This map was drawn up by M. Poirson, an able engineer of Paris, from materials furnished to him by M. Oltmanns and me. It embraces the immense extent comprehended between the 15° and 42° of latitude, and the 75° and 130° of longitude. At first I meant to extend this map to the south as far as the mouth of the Rio San Juan, for the sake of indicating different canals, of which the construction was proposed to the court of Madrid, and which would serve to establish the communication between the two seas, to be discussed in the second

chapter of this work. But on perceiving that the peninsula of Yucatan, and the coast of Monterey, would not be represented with the developement which they deserved, I chose rather to preserve a larger scale, and to extend my map no farther south than the gulf of Honduras.

The principal part, that which comprehends the kingdom of New Spain, is a faithful copy of my large map, of which I have given an analysis. The Yucatan was added from the map of the gulf of Mexico, published by the Deposito Hydrografico of Madrid. New California was taken from the atlas which accompanies the account of the voyage of the corvettes Sutil and Mexicana, and from a memoir of M. Espinosa, printed in 1806, entitled, Memoria sobre las observaciones astronomicas que han servido de fundamento a las cartas de la costa N. O. de America, publicadas por la direccion de trabajos hidrograficos. When this memoir gave different results from those contained in the Relacion del viage a Fuca, they were preferred as founded on more solid bases*. The

* I have placed Monterey in latitude 36° 35′ 45′′, and longitude 124° 12′ 23′′, and Cape S. Lucas in latitude 22° 52′ 33", longitude 112° 14' 30". The longitude of Monterey which I have definitively adopted with M. Espinosa, differs less from that of Vancouver than the result published by M Antillon. The difference between the opinion of the Spanish navigator and that of the English navigator is only an arc of 18' as already stated. (Here it is of importance to observe, that the commencement of this geographical introduction,

work of M. Espinosa served me also for the small group of islands, named by M. Collnett the archipelego of Revillagigedo, in honour of a Mexican viceroy.

The islands of San Benedicto, Socorro, Rocca partida and Santa Rosa, situated between the 18° and 20° of latitude, were discovered by the Spanish navigators in the commencement of the sixteenth century. Hernando de Grixalva discovered in 1533 the island of Santo Tomas, now named Isle del Socorro. In 1542, Ruy Lopez de Villalobos landed on a small island, to which he gave the name of la Nublada. He indicated very well its true distance from the island of Santa Tomas. This Nublada of Villalobos is now called San Benedicto. It is not so certain that the Rocca partida of the same navigator is the island of Santa Rosa of the modern hydrographers, for the greatest confusion prevails as to the position of this rock. Juan Gaetan places it even two hundred leagues to the west of the island of Santa Tomas.

This last island is marked at 19° 45' of latitude, and as a shallow of thirty-six miles in length, on the map of Domingo de Castillo drawn up in

from p. i. to p. xxxiii. was composed at Berlin in the month of September 1807, and that the remainder was published in the spring of 1809).

* Ramusio, t. I. p. 375 (edition of Venice, 1613).

1541, and found in the archives of the family of Cortez at Mexico. Since that time the groupe of islands of Revillagigedo has only been thrice seen; namely, by the pilot Don Josef Camacho, in 1779, in a navigation from San Blas to New California; by captain Don Alonzo de Torres, in 1792, in a voyage from Acapulco to San Blas; and lastly, by M. Collnett* in 1793. The observations of these three navigators are extremely discordant. Yet it would appear that M. Collnett has fixed exactly enough the position of the Isle del Socorro, from several series of distances of the moon from the sun. It is from these same distances calculated by Mason's tables that the whole groupe of islands has been thrown too far east.

As to the countries conterminous with New Spain, we have used for Louisiana the fine map of the engineer Lafond; and for the United States the map of Arrowsmith, rectified from the observations of Rittenhouse, Ferrer, and Ellicott. The positions of New York and Lancaster were dis

* Collnett's voyage to the South Sea, p. 107. M. Collnett finds the latitude of Cape San Lucas 22° 45', and the longitude 112° 20′ 15′′. This latitude appears to be nearly seven minutes false! The mountain of San Lazaro, whose position is fixed by M. Collnett at 25° 15′ of latitude, and 114° 40′ 15′′ (p. 92 and 94.) is undoubtedly not the same as that which Ulloa called, in 1539, Cape of San Abad, and which I have placed (after M. Espinosa) in 24° 47' latitude, and 114° 42' 30" longitude.