290 LAKE SUPERIOR. 



about equal, and there are still others in which the individuals of 

 the same species are alternately at different seasons of the year either 

 marine or fluviatile ; this is particularly the case with such as ascend 

 from the sea into the rivers at the spawning season, to deposit their 

 eggs in waters more genial to the growth of their young than those 

 in which they are mainly to live when full-grown. 



Percoids belong to those families of which there are certain pro- 

 portions of strictly freshwater, and certain proportions of strictly 

 marine genera, the number of marine species being however much 

 greater than that of the freshwater ones, and very few of the species 

 having the power of enduring both the freshwater and the sea. 



That the family of Percoids, as it is now circumscribed, is in 

 the main a most natural group, cannot be doubted, especially if 

 we remove from it such genera as Trachinus, Uranoscopus, Sphy- 

 raena and a few others ; there remains however a question, not to be 

 decided here, how far Sparoids and Sciaenoids should be considered 

 as distinct. Indeed, at different times, in two editions of the same 

 work, Cuvier in his Animal Kingdom has successively associated 

 them in one great family, and divided them into two distinct groups. 

 The fact is that these fishes are closely related, and it is for future 

 investigations to determine the value of those characters upon 

 which the distinction rests, which consists only in the serrature of the 

 opercular apparatus, the presence or absence of teeth upon the 

 palatine bones, and the degree of development of the so-called 

 mucous canals in the head, characters which have not even been 

 strictly adhered to in the arrangement of individual genera. 



Whatever may be their closer or more remote affinities, the 

 Percoids of the Canadian lakes, as well as those of the other fresh 

 waters of North America, are much more diversified than those of 

 the freshwaters under similar latitudes in the Old World. This 

 is not the case with Lake Superior itself, for, on the contrary, that 

 lake furnishes but few true Percoids ; but the other great lakes teem 

 with a variety of genera and species of that family, which among 

 themselves, as well as with reference to the common type of the 

 whole family, differ much more from the Percoids than those of 

 Europe ; I need only mention the genera Pomotis, Centrarchus, 



