8 PROCEEDINGS OF THE AMERICAN ACADEMY 



branclms from Portage Lake, near Lake Superior. He had 

 had them in his possession since December, 1856, and from 

 that time until the succeeding June they took no food. At 

 the present time they devour earth-worms greedily. During 

 the past winter, the water in which, they were kept had re- 

 peatedly been frozen solid. They were pumped up from the 

 lake, and it is only during the winter season that specimens 

 are obtained. 



Professor Horsford gave the results of several experiments 

 by his pupils to determine the commercial value of saltpetre 

 by a new method proposed by himself some time since, and 

 then announced to the Academy. The results corresponded 

 very accurately with each other, and with those obtained by 

 more elaborate chemical processes. 



Four buudred aiid forty- Aftb meeting. 



December 8, 1857. — Monthly Meeting. 



The President in the chair. 



The Corresponding Secretary read a letter from Professor 

 Edward Robinson, accepting the Fellowship of the Academy. 



Professor Agassiz spoke of the various existing systems 

 of Classification of Fishes, characterizing them all as incom- 

 plete and artificial. He analyzed those of anatomists and 

 zoologists, showing how each failed to conform to the natural 

 system. He referred to his own proposed classification, based 

 on the characters of the scales, and said that he had given 

 that up also as too artificial. The true indications, he 

 thought, were to be found in the embryonic development of 

 this class of animals. Development without an amnios or 

 allantois is common, he said, to naked reptiles and fishes. 

 The scaly reptiles, birds, and mammals are associated by 

 their circulation as a natural group. Among fishes he rec- 

 ognized four groups on a par with the natural divisions of 

 amphibians. He proposed the name of Salachians, to include 

 the sharks and skates, which differ in the structure of their 



