74 



mus larvai and Entomostraca — eighty-nine per cent, anrt eleven per cent. 



respectively — while the larger had tai^on only aquatic larva? — nearly all 

 ephemerids. 



Finally, eight of the slender, active and wholly prcdaceous little hrook 

 silversides (LdbidestJirs sknihi.^) had eaten a single fish, fourteen per cent, of 

 Entomostraca. and jihoul eighty per cent, of insects — somewhat more than 

 half of a(iuatic origin. In brief, tlie -structures of Lahidesllies. the habits 

 of Phenacobius and the darters, and the differences in size of the species 

 of Boleosoma and lladropterus were all reflected in the food of this little 

 group. 



The obverse fact of the unifying effect of similarity of alimentary struc- 

 tures is apparently shown i)y a small collection of minnows, all itelonging 

 to the first two groups of the paper cited al>ove.* made from an extremely 

 muddy little creek in .Jersey county, wiiich contained no visible vegetation 

 and few, if any, Entomostraca. Twelve of these fishes, representing the 

 genera Campostoma, Pimephales, llyliorhyncluis. Ilybognathus and Note- 

 migonus, agreed in food almost precisely, all having swallowed the fine 

 mud of the creek bottom, with a slightly varying admixture of unicellular 

 AlgiB and vegetable debris. 



As an example of a contrast between two species agreeing in alimentary 

 structures, but differing in size and somewhat, also, in habitual range, we 

 may take three exaujples of Noh-opis hctrrodon and three of yolropis mrgn- 

 loi)s, captured at McIIenry. May s. i.sso. More tiian half the food of the 

 latter group consisted of vegetation, and of the former only ten per cent. 

 The remaining ninety per cent, of the food of hrkrodnn was Entomostraca; 

 but these were not r'epresented at all in the ntrfjaJoiis. the remaining food 

 of these specimens consisting of insects and am]jhipo(l Crustacea. 



Sensible and even conspicuous dilTerences in food often appear between 

 groups which are neither widely separate in classitication nor yet distin- 

 guished by marked differences in alimentary structures, as between species 

 of the same genus. Sometimes these are apparently due to differences in 

 habit with respect to the search for food, but sometimes seem dependent 

 upon distinction of habit or preferences even mc)re obscure. 



Six specimens of the channel cat (Id(drii!< pioictatu.-o. taken at IVoria, 

 October 6, 1887, had eaten insects, mollusks and vegetation at the rate of 

 forty-one, nineteen and forty i)ercent., res]H>ct ively. tlie vegetation being 

 nearly all Cladopliora and I'ot;iiiiogeton, while the same number of bull- 

 heads (Audunis nchidosHs) had derived tliirty-seven i)er cent, of their food 

 from insects and sixty-three pei' cent, from mollusks. The difference here 

 was substantially a largei' ratio of mollusks for Amiurus, reiilacing the 

 vegetable food of the Ictalrus group. By a comparison of these dilTer- 

 ences with those detected between the species at large, as explained on 

 pages 45(>-4(il, it will be seen tliat the former do not represent the specillc 

 differences in food, but simply give evidence that the two species may be 

 differently affected by the same conditions. 



Other specific differences in the same genus are shown by the collections 

 made Oct. 27. 1875, from Peoria lake. Eight examjiles of the wall-eyed 

 pike (Stiznstedion rilrcum) had (>aten only soft-tinned tislies. — excepting one 

 small sunttsh, — while four of t(Mi specinuMis of the related species 5". 0/>ja- 

 dense, had eaten spiny-tlniKHl tislies, and in only three were tlie tishes rec- 

 ognizable as belonging to the soft-linned species. Three specimens of Mi- 

 cropetrus taken witii the above had eaten cray-tishes and tishes (including 

 a cattish). 



Among my s]ieciniens of the sucker family (Catostomatidae). a lot ol>- 

 tained at Qiiincy. .\ug. 25. 1887, are comparable for tlie present purpose. 

 Four examples each of IrHobux itrux and I. cupvimlJa ]iresented a decided 

 contrast with respect to the elements of their food, that of /. »>•»,•< consist- 

 ing almost wholly of Chironomus larva*, with large (juantities of dirt, while 

 three of the specimens of I. cypnnclki had eaten scarcely anything but Al- 



*See the preceding page. 



