the movement of the waters at the same season and with the same wind. Why is the Gulf-stream sometimes borne on the coast of Florida, sometimes on the border of the shoal of Bahama? Why do the waters flow, for the space of whole weeks, from the Havannah to Matanzas, and (to cite an example of the corriente por arriba, which is sometimes observed in the most eastern part of the main land during the prevalence of gentle winds) from La Guayra to Cape Codera and Cumana? As we advanced, on the 25th of March, towards the coast of Darien, the north-east wind increased with violence. We might have imagined ourselves transported to another climate. The sea became very rough during the night yet the temperature of the water kept up (from latitude 10 degrees 30 minutes, to 9 degrees 47 minutes) at 25.8 degrees. We perceived at sunrise a part of the archipelago* of Saint Bernard, which closes the gulf of Morrosquillo on the north. (* It is composed of the islands Mucara, Ceycen, Maravilla, Tintipan, Panda, Palma, Mangles, and Salamanquilla, which rise little above the sea. Several of them have the form of a bastion. There are two passages in the middle of this archipelago, from seventeen to twenty fathoms. Large vessels can pass between the Isla Panda and Tintipan, and between the Isla de Mangles and Palma.) A clear spot between the clouds enabled me to take the horary angles. The chronometer, at the little island of Mucara, gave longitude 78 degrees 13 minutes 54 seconds. We passed on the southern extremity of the Placer de San Bernardo. The waters were milky, although a sounding of twenty-five fathoms did not indicate the bottom; the cooling of the water was not felt, doubtless owing to the rapidity of the current. Above the archipelago of Saint Bernard and Cape Boqueron we saw in the distance the mountains of Tigua. The stormy weather and the difficulty of going up against the wind induced the captain of our frail vessel to seek shelter in the Rio Sinu, or rather, near the Punta del Zapote, situated on the eastern bank of the Ensenada de Cispata, into which flows the river Sinu or the Zenu of the early Conquistadores. It rained with violence, and I availed myself of that occasion to measure the temperature of the rain-water: it was 26.3 degrees, while the thermometer in the air kept up, in a place where the bulb was not wet, at 24.8 degrees. This result differed much from that we had obtained at Cumana, where the rain-water was often a degree colder than the air.* (* As, within the tropics, it takes but little time to collect some inches of water in a vase having a wide opening, and narrowing towards the bottom, I do not think there can be any error in the observation, when the heat of the rain-water differs from that of the air. If the heat of the rain-water be less than that of the air it may be presumed that only a part of the total effect is observed. I often found at Mexico at the end of June, the rain at 19.2 or 19.4 degrees,