316 Annals of the Carnegie Museum. 



1 6. Zacco pachycephalus. Zacco temmincki from Japan. 



17. Acheilogiuithus himantegiis. Acheilognathus cyanostigma from 



Japan. 



18. Rhinogobiiis formosaniis. Rhinogohius hadropterus from Japan. 



Not only such a relationship is manifested by the peculiar species, 

 but by the rest of the fresh-water fishes as well. Of the twenty-one 

 species, which occur outside of Formosa, Clarias fuscus, Capoeta 

 semifasciolata, DistcecJiodoii tumirostris, Culter hrevicauda, Cultriculus 

 kneri, Ophicephalus tadianus, and Ophicephalus maculatns are species 

 of southern affinities, because they are distributed in Indo-China and 

 South China, but not to the north of the Yang-tze-kiang. Polyacan- 

 tJiiis is a genus of the Indo-Malayan type, extending into the Malay 

 Archipelago, but not occuring in eastern Asia. Such being the case, 

 Polyacanthus operculatus may be included in this category, though it 

 has been recorded from Tien-tsin, North China. 



Eleven species are of more or less general distribution, extending 

 from South China to Corea and Japan proper, and one occurs only 

 upon the Chinese mainland. Finally, Sicyoptcnis japonicus is one 

 of the peculiar species of Japan, and its relationship is somewhat 

 dubious, as it is not known to occur on the Asiatic continent. 



It will thus be seen that all the Formosan fresh- water fishes which 

 have Chinese affinities ditTerentiate into more or less distinct species, 

 while those of southern affinities have remained unchanged. More- 

 over, eight per cent of the non-peculiar species have been recorded 

 from the Chinese mainland. These two facts explain very clearly 

 that the island had been preoccupied by the fresh-water fishes of the 

 Chinese fauna when those of the southern affinities appeared through 

 South China. 



Next to the prevalence of Chinese affinities, the total absence of 

 any indication of affinity to the fresh-water fish fauna of the Philippine 

 Islands and Malay Archipelago is a very striking fact. As shown in 

 the table, only one species is recorded from the Philippines, namely 

 Ophicephalus maculatns. However, as it also occurs in South China, 

 its way of dispersal is clearly indicated, though there is no record of it 

 in India and the Malay Acrhipelago. Cyprinus carpio is another 

 species which has been recorded from Java. But it is evident that 

 the carp is not a native of Java, but an introduced species. 



According to Leonhard Stejneger, there exists the same relationship 

 between Formosa and the Philippine Islands with regard to the her- 



