^'l89i!''] PROCEEDINGS OP THE NATIONAL MUSEUM. 19 



The exi)lauatioa is probably conuected with the age of any given 

 geuus. Those genera with many species and wide distribution are 

 evidently now at their prime, while tbose with wide distribution and 

 few species, occupying isolate places are probably remnants from 

 another age, and genera with few species and narrow distribution are 

 very probably later diftereutiations. There are, of course, cases which 

 will not be classified thus. Callichthys and Hoplostermim are cases in 

 hand which have already been mentioned. Flatystoma (List iii) otfers 

 another instance, being composed of a single species distributed over 

 nearly the whole of the region east of the Andes and north of Buenos 

 Ayres. 



Another fact worthy of mention, though not directly illustrated by 

 these lists, is that the species of wide distribution belong to genera of 

 many species and wide distribution. Genera of many species frequently 

 have one or more species of wide distribution. On the other hand 

 genera of few species and narrow distribution usually have species of 

 restricted distribution. 



The variability of species of wide distribution has already been men- 

 tioned. 



We shall now take up the zoogeography more in detail. Too great 

 stress must not be placed on our present knowledge; the details of the 

 distribution of not one species is as yet worked out. The absence of 

 certain genera from the Rio Magdalena and the Eio Plata is probably 

 due to our lack of knowledge. The general results, however, will per- 

 haps not vary greatly from what may be deduced from the present data. 



A word as to the preparation of the lists. The entire catalogue was 

 read and tlie genera (exclusive of marine) having similar geographical 

 boundaries were placed together, the result obtained being presented 

 in the foregoing lists. The regions covered by each list are, therefore, 

 the necessary outcome of the facts. There are, naturally, a number of 

 genera which can not be placed in any of the lists. 



The first list gives the genera characterizing the Fuegian subregion 

 of the southern zone. Although a few genera {Chirodon, Pygidium) 

 have representatives here, its fauna is such as to separate it very dis- 

 tinctly from the neotropical realm and it is included here more for con- 

 venience than for its afliuity with the rest of South America. 



The second list, and the third and fourth with the exception of those 

 genera found also in the Mexican subregion and so marked, present 

 the genera which characterize the Brazilian subregion as a whole. A 

 few of the genera have not been found in the Rio Plata and the Rio 

 Magdalena. 



The fifth list characterizes what may be termed the Pacific province 

 of the Brazilian region. It includes the territory west of the Andes, 

 between Costa Rica* and Peru. 



* The Rio Chagres certainly does not belong to the Mexican subregion. 



