industrious have an opportunity of making a little money independently. At night they all present their tickets and are paid on the spot for any extra work. From the harvesting-ground we followed the carts down to the place where their burden is deposited. On their return from the plantation the negroes divide the day’s harvest, and dispose it in little mounds on the drying-ground. When pretty equally dried, the coffee is spread out in thin even layers over the whole enclosure, where it is baked for the last time. It is then hulled by a very simple machine in use on almost all the fazendas, and the process is complete. At noon we bade good by to our kind hosts, and started for Juiz de Fora. Our stage was not a bad imitation of Noah’s ark, for we carried with us the beasts of the field and the birds of the air and the fishes from the waters,[41] to say nothing of the trees from the forest. The party with whom we had passed such pleasant days collected to bid us farewell, and followed us, as we passed out from the gate, with vivas and waving hats and handkerchiefs. The following day we were fortunate in having cool weather with a somewhat cloudy sky, so that our ride of ten hours from Juiz de Fora to Petropolis, on the top of the stage, was delightful. The next morning in driving down the Serra to Mauá we witnessed a singular phenomenon, common enough, I suppose, to those who live in high regions. As we turned the corner of the road which first brings us in sight of the magnificent view below the Serra, there was a general exclamation of surprise and admiration. The valley and harbor, quite out to the sea, were changed to a field of snow, white, soft, and fleecy, as if fallen that night. The illusion was perfect, and though recognized at once as simply an effect of the heavy morning fog, we could hardly believe that it would disperse at our approach and not prove to be the thing it seemed. Here and there the summit of a hill pierced through it like an island, making the deception more complete. The incident was especially interesting to us as connecting itself with our late discussions as to the possible former existence of glaciers in this region. In his lecture a few nights before, describing the greater extension of the ice in former geological ages, when the whole plain of Switzerland between the Alps and Jura must have been filled with glaciers, Mr. Agassiz had said “there is a phenomenon not uncommon in the autumn in Switzerland which may help us to reconstruct this wonderful picture. Sometimes in a September morning the whole plain of Switzerland is filled with vapor which, when its pure white, undulating surface is seen from the higher summits of the Jura, looks like a snowy ‘mer de glace,’ appearing to descend from the peaks of the Alps and extending toward the Jura, while from all the tributary valleys similar masses pour down to meet it.” It was as if the valley and harbor of Rio had meant to offer us a similar