cellular substance which is full of water; these, which are evidently intended by nature to supply the vegetation of the tree during the dry season, are often dug out by travellers for the sake of the water they contain, each tuber yielding about a pint of excellent quality. Another wild fruit common here, as it is also about Crato and Pernambuco, is produced by a tree growing to the height of from thirty to forty feet, called Pitombeira; it is the _Sapindus esculentus_, St. Hil.; the fruit is produced in large bunches, resembling in size the common grape; the outer covering is hard, but the embryo, or kernel, is covered with a thin transparent, sweetish, acid pulp, which alone is eaten; the kernels are said to poison turkeys if eaten by them. Next day, the twentieth of January, we left Cachoeira, early in the morning, and after travelling about five leagues we arrived at a Fazenda, called Marmeleira, where we rested during the middle of the day. Shortly after leaving Cachoeira we began to ascend a small Serra, by a very gradual inclination, but owing to its rocky nature the road was very bad in many places. This range consists of gneiss, the stratification of which is nearly vertical; large blocks of the same kind of rock were very commonly seen along the road, but on the ascent of a small hill I observed many round blocks of a coarse whitish-sandstone. On the journey we crossed several rivers, which although very small, or wholly dried up in the arid season, were now so much swollen by the recent rains, that we could with difficulty pass some of them. After crossing the Serra we entered a fine valley about a league in length, well wooded with large trees, and having a small river running through the centre of it; the end to the westward was pretty well cleared of trees, and there is a little hamlet called Rosario, beyond which we had nearly half a league to travel before reaching our resting-place. On this journey were found several new species of the beautiful genus _Angelonia_, one of the most remarkable being a fine climbing shrubby species (_A. bracteata_, Benth.); on dry rocky places I observed several kinds of _Cactus_, which are not so common on the western side as they are on the east of the Serra de Ibiapaba. The genus _Loasa_, so very common in Chili and Peru, has only few representatives in Brazil; one which I have called _Loasa rupestris_, was found here in rocky places by the side of a stream just before we reached our resting-place; like other species of this genus, the whole plant stings very powerfully, and hence, in common with other stinging plants, it is called by the inhabitants Cançançao. As the sun shone out brightly during our stay at Marmeleira, we were enabled to dry all our clothes and other things that had been soaked by the rains. Here I prescribed for the mistress of the house, who was in a very delicate state of health; in return for this we were abundantly supplied with milk, which at this season is plentiful.