181 



COMPARISON OF THE GROUPS. 



It will 1)6 remembered tliat'the groups were I)a8ed upon differ- 

 ences in the structures relating to the appropriation and mastica- 

 tion of food. It is consequently from a comparison of the ratios 

 of these groups that we shall derive the most interesting facts 

 relating to the correspondence'^ of food and structure. The most 

 conspicuous result is the great preponderance of mud in the 

 intestines of the fishes of the first group, characterized by an extra- 

 ordinarily elongate intestine, and by pharyngeal teeth destitute 

 of hooks and provided with a broad grinding surface. Here, as 

 already noted, mud, sand, and gravel amounted to about three 

 fourths of the matter ingested, while in the third and fourth groups 

 only trivial and accidental quantities occurred. In the second 

 group, on the other hand, with intestines intermediate in length, 

 mud was still abundant, but much less so than in the first, aver- 

 aging less than half the whole. If we exclude this indigestible 

 matter, however, we shall find the first group still further distin- 

 guished by the predominance of vegetation as compared with 

 animal matter, the latter being only about one third the former, 

 while in Groups III and IV, on the other hand, vegetation 

 amounts to about one third the animal food. The groups last 

 mentioned, distinguished from each other as they are, only by the 

 presence of a masticatory surface on the pharyngeal teeth in the 

 first, and its absence in the second, differ scarcely at all in their 

 general food characters, and this structural feature seems therefore 

 to be of little significance. In both the animal ratio amounts to 

 seventy-five per cent., and vegetation stands in each at twenty- 

 five ; while insects are respectively fifty and sixty-one. It is true 

 that we find neuropterous larvas greatly predominant in the first 

 group, making one fourth of their food, and Chironomus larvaB 

 in the second amounting to sixteen per cent. The second of 

 these facts we find upon analysis to be evidently due to Phenaco- 

 bius, by which genus nearly all the Chironomus larvte were taken ; 

 and this, as already shown, is explained not by any structural 

 feature, but by its peculiar habitat; and when we note that aquatic 

 larvae together amount in Group III to twenty-five per cent., and 

 in Group IV to twenty-seven, we see that the significance of the 

 difference mentioned disappears. A similar explanation is found of 

 the difference in the ratios of Entomostraca, — thit of the fir^t group 

 amounting to twenty per cent, and that of the second only to four. 

 An examination of the tables shows that this predominance in the 

 group first mentioned is nearly all traceable to Hemitremia. a ver> 

 small fish with rather elongate gill-rakers. 



The importance of these gill structures is still more clearly 

 indicated, as already noticed, by the difference between Notemi- 

 gonus and Chrosomus of the second group, and clearly far out- 

 weighs the structure of the teeth as an indication of the food 

 habits of the fish. 



The general conclusions reached may be thus briefly stated : 

 An extraordinarily elongate intestine indicates the limophagous 

 habit, rather than an especial preference for vegetable food. Tho 



