OF ARTS AND SCIENCES. 195 



pala30zoic period, was next treated at length by Professor 

 Agassiz. 



In reference to Professor Agassiz's communication on the 

 paleozoic animals, all of them aquatic, Dr. Pickering re- 

 marked, that the oceanic animals of the present day are 

 almost as universally diftused as were those of paleozoic 

 times. At least, a division into tropical and non-tropical 

 will require the oceanic tropic to be placed full fifty degrees 

 from the equator : there will then remain, outside of this 

 belt of a hundred degrees, a few oceanic species that may 

 be regarded as non-tropical. 



Dr. A. A. Gould mentioned the fact that the land shells of 

 the Sandwich Islands are peculiar in being distinct in their 

 characters for each island : they are species of the genus 

 Achatinella. The island of Metia has three species peculiar 

 to it. The shells of the Feejee Islands do not resemble those 

 of the Navigator's Islands, but are like those of the Society 

 Islands, twenty degrees distant ; a fact which Dr. Gould con- 

 sidered evidence of the former existence of an extensive con- 

 tinent now submerged. 



Dr. Pickering said he did not think there was knowledge 

 enough as yet of the shells of the Society Islands in the pos- 

 session of naturalists to assert positively that the Metia shells 

 do not exist on them. If this be really the case, it would be 

 evidence of comparatively modern creation, as the island of 

 Metia is a coral-island. 



Professor Agassiz mentioned that even the fishes of differ- 

 ent islands of the Sandwich group differ from each other, as 

 had been proved by specimens in his possession. 



Professor Gray, in a brief rejoinder to the remarks of Pro- 

 fessor Agassiz upon his communication at a former meeting 

 (p. 177), suggested that an obvious reason why the terrestrial 

 species of the arctic regions were so generally alike round the 

 w^orld, while those of the temperate zone were less so, and 

 those of the tropics mostly different, might be found in the 

 geographical position of the regions in question, — the masses 

 of land in the former being essentially contiguous, in the lat- 



