1899.] OBNIS OF SAO PAULO. 509 



Evidence of this is given by the fact that Pelzeln has not 

 meutioned the hne which coincides with the boundary between the 

 States of Parana and Sao Paulo. Although some of the more 

 striking pieces of evidence of this line have been given only in mj'- 

 paper, others are due to batterer's collections. This line marks the 

 northern limits of a number of Argentine species which occur in 

 the three southernmost States of Brazil, but do not extend north 

 of it into Sao Paulo and Rio. Besides certain characteristic 

 species, such genera as Ci/anotis, Phloeocryjjtes, Anumbius, Gliban- 

 orni^ and Haplospiza characterize this " ^nwrnSiMS-line," as it may 

 be named. 



I cannot forbear to mention that I have been astonished to find 

 that such results as those I have arrived at on the faunistic 

 boundary-lines in Eio Grande do Sul should have beeii disregarded 

 by zoologists. However, I shall continue to work on in the same 

 manner, and if with this help there cannot be constructed a 

 complete system, we shall obtain at any rate exact data for the 

 analysis of the faunas of some of the States ; and if the same task 

 be undertaken in other States of Brazil, the results must without 

 doubt be satisfactory — as a piece of mosaic-work, but a definite one. 

 It is evident that such work can only be the result of extensive and 

 exact ex]:)lorations of restricted areas, and more of it is to be 

 expected from Museums than from observers, who spend but a 

 short time in one country. 1 hope that Dr. Groeldi, continuing his 

 work in the Para Museum, will investigate the State of Para in the 

 same manner. 



In concluding these general remarks, I wish to state, as the 

 result of my studies, that South-eastern Brazil, from Rio Grande 

 do Sul to Bahia, and probably farther northward, forms a natural 

 province of the Neotropical Region, which contains t\^o Sub- 

 divisions (see Map, Plate XXVII.). One of these extends from 

 Rio Grande do Sul to Rio de Janeiro ; the other from the North 

 down to Sao Paulo. The northern Subdivision extends along the 

 coast of Sao Paido to Iguape and probably farther southward, but 

 is here restricted to a narrow coast-zone. This is separated by a 

 narrow chain of mountains from the highlands, and these high- 

 lands towards the west pass into campos, which have the same 

 fauna as the campos of Minas, Goyaz, and Matto Grosso. We 

 have, therefore, in Sao Paulo three faunistic subdivisions, repre- 

 senting from the west to the coast successively the central, 

 marginal, and littoral fauna. The two last are separated by the 

 Serra do Mar, which is only a few miles broad, but supplies a 

 difference of altitude of more than 700 metres and a difference of 

 temperature of 3° C. or more. This is the reason why many Bahia 

 species which do not occur in the intei-ior of Sao Paulo are found 

 along the coast. 



I now proceed to the special discussion of my paper and its 

 zoo-geographical results. 



There are two new species described in my paper — Chrysotis 

 schmidti, closely allied to Oh. awipalliata, but with the bend of 



