The work reported here was carried-out on the invitation of Dr. Henry Kumm, Director of the Rockefeller Foundation, and by appointment from Dr. Henrique Aragão, Director of the Instituto Oswaldo Cruz. It was done during the investigation of sylvan yellow fever, in June 1947, with a view to establishing the phyto-ecological conditions of the county of Passos. The pe¬riod was, however, too short for definite conclusions to be reached. Thanks are due to Dr. O. R. Causey, Chief of Research on Yellow Fever for transpor¬tation and other help. THE REGIONAL VEGETATION. Aerial photographs of the county of Passos shoto that it is covered by three great types of vegetation: Rain Forest, Secondary Pasture Land and Scrub.1 Detailed investigation, however, brings out the fact that these correspond to different seres; furthermore, each type presents not only the specific, characteristics of the biological form dominant for the climate, but also are at various stages, which express HABITATS differing from those of the normal sere. The phytogeographic survey of the region shows that most of it is now covered by secondary pasture land (disclimax) in which Melinis minutiflora, v. "fat grass" (fig. 1), predominates. The mosaic of Rain Forest and of small patches of Scrub reveals the effects of human intervention (BARRETO, H. L. de Mello 1); consequently, all the formations have to be regarded as secon¬dary, though some of them probably include relicts of the primitive climax (WARMING, E. 2). On close examination, the Scrub cannot be considered as the climax, because of the following facts: 1. In the zone of Rain-Forest stretches of forest are present in very varied topographic conditions and the reconstitution of the associations show that man has destroyed an ecological unit (fig. 2). 2. In the zone of Scrub the characteristic patches are small. The banks of rivers and brooks, the valleys and ravine and whatever the soil has retained some humidity, is being invaded fry Rain Forest, which seems to be growing under optimum conditions. The Scrub is thus limited to small belts on the calcareous mountains and on sandy soils with alcaline depths (pH abo¬ve 7) which do not retain enough moisture for the Rain Forest that is progres¬sively restricting the area occupied by Scrub. In view of the topographic and present climatic conditions the Rain Forest must consequently be regarded as the regional climax. The presence of ecologically contradictory elements and associations shows that the real problem is that of the fluctuations of the climate of Passos or even of Minas Geraes during the quaternary and recent periods (DAN-SEREAU, P. : 3), a subject on which little is known and which is tied to the evolution of the climate of Brazil (OLIVEIRA, E. : 4) . The transformation of Scrub into Rain Forest has been - observed by the author before, in other parts of Brazil (VELOSO, PL P.: 5) . It seems probable that the Rio Grande has also greatly influenced the change of the regional vegetation, by invading areas of Scrub and dislocating the limit of the Pluvial climate towards the Canastra Range, though there are remnants of Scrub (postclimax) transfor¬med into secondary open country (disclimax, fig. 5) by human devastation and the setting of fire to the land. VEGETATION GROUPS OF THE PLUVIAL TYPE. The map of the region also shows that at the present time the small patches of forest (whether devasted or intact) occupy the least accessible places, such as valleys, peaks and abrupt slopes (fig. 2). Even these are now being destroyed, so that in the near future this forested region will be en¬tirely reduced to poor pasture land unless energetic measures of conservation are undertaken in time. The Special Service for Prophylaxis against Yellow Fever installed two of their four Stations for the Capture of Mosquitos in this area, one of them at Batatal and the other at Cachoeira, which have separate formations each of them composed of several associations. Other vegetation formations were also analysed, from the synecological point of view, so as to ascertain of which degree of succession their associations belong. These phytosociological sur¬veys give an idea of the principal characteristics of each station. BATATAL FORMATION. The abrupt nature of the valley has rendered this location inappropriate for agricultural purposes since colonial times. The relict of the primitive forest climax saved by this circumstance has expanded gradually to zones whose paedologic conditions favour the eatablishment of mesophilous species. The aerial photograph shows two small stretches of forest, one apparently primi¬tive, the other composed of associations belonging to the subclimax of the subsere. CACHOEIRA FORMATION. Aerial photographs show that this station is crossed by a small river, which divides it into two separate parts. The first, which presents ecological conditions similar, though not identical to those of Batatal, is favoured by topography and apparently remains primitive forest. Though the topography of the other, on the whole, favours the establishment of groups belonging to the normal sere of the climax, is has been partly devastated recently and the aspect of the associations has been completely modified. It was is this part that the four posts for the capturing of mosquitos were set up. The first forest is favoured by deposition of organic matter, washed out from the nearby devasted areas by torrential rains, and thus provides, an appropriate HABITAT for the climax species with certain hygrophilous trends of the ecological quasiclimax type. This association seems to have reached a biological equilibrium, as the dominates. Gallesia gorarema and Cariniana legalis (fig. 10), present an optimum vitality with a vigorous habit and a normal evolutionary cycle. The Cariniantum legalis Gallesiosum equilibrium, corresponds however, to a provisory association, because if the moving of soil by torrential rains should cease it would become possible…