48 BULLETIN III-' THK BUREAU OF FISHERIES. 



/'A(Y/^'/,/w.— Tlircc stoiiuu'hs coiitiuncd 10, 5, and 'l per cent adult stone tlii's 

 and 12 from 3 to 25 per cent stone-Hy nymphs. 



Orthoptem. — The .stomach contents of 4 trout consisted of 50, 6, 5, and 2 po 

 cent pieces of grasshoppers (locusts). 



IIeiiii]>tera. — Five stomachs contained 0.25 to 2 per cent Hemiptera. 



Trii}i<ii>fi'r((. — The contents of 5 stomachs consisted of 40, is, 10,5, and 2 jki- 

 cent adult caddis flies; 12 contained from 0.25 to 15 per cent caddis-fl}' larvjv. luid 

 14 contained larval cases varying in amount from 1 to 25 per cent. 



Lejndop.em. — Eight per cent of the contents of 1 stomach consisted of remains 

 of moths. 



Chironoiiiidx. — Eight specimens had eaten from 0.5 to S per cent ciiironomid 

 larva?. 



Coleoptera. — Beetles constituted 1 to 3 per cent of the stomach contents of 7 

 specimens, and larvse of diving beetles (water tigei's) constituted from 3 to 8 per cent 

 of the contents of 7 specimens, 5 of which contained no other Coleoptera. 



Ilijmenoptera. — Ants constituted from 1 to 20 per cent of the stomach contents 

 of 13 specimens. 



Inject frmjments. — From 1 to 88 per cent of the contents of all the stomachs con- 

 sisted of small fragments of insects. 



Yegetitble matter. — Two stomachs contained 8 per cent vegetable matte)-, one 5 

 per cent, two 1 per cent, and one 0.5 per cent. 



Sand. — The stomach contents of 9 individuals consisted of 1 to 60 per cent sand, 

 whicli was deriv-ed chiefly from the cases of insect larva3 and pupa?. 



The average percentages for these 18 golden trout are as follows: Arachnida 

 0.2'.>, P^i^henierida (May-fly nymphs) 7, Plecoptera (adult stone flies and nymphs) 12.2, 

 Ortlioptcra :;..">. Hemiptera 0.27, adult and larval Trichoptera 7.22, cases of larval 

 Trichopti'ra 7, Lepidoptera (moths) 0.45, Diptera().9, chironomid larvae 1.5, Coleoptera 

 2.67, Hymenoptera (ants) 4.1, insect fragments 41.9, vegetable matter 1.3, sand 9.7. 

 These percentages show that the most important elements of the food of the golden 

 trout at the time these s])e(unens were caught were May-fly nymphs, adult stone 

 flies and nymphs, and adult cadtlis-tlies and larvie. Ants ranked next in importance. 

 Only two of these stouuieh> cuutuined threadworms (Nematoda). 



DLSCUSSION OF 1{ESULTS. 



If the foregoing results may be taken as representative of the various streams 

 from which these specimens were obtained, the trout in them were dependent almost 

 wholly on insect life for their food at the time these investigations were made. 

 In fact, judging from the sc^arcitj^ of other animal forms that might serve as food, 

 adult insects and the atjuatic larvte of insects must play a very important role in the 

 food supply of these trout during the entire year. It is possible, of course, that the 

 small trout maj^ be eaten by the larger ones at times, but no such cannibalistic 

 tendencies were found among the specimens examined, the only indication of canni- 

 balism noted being the single trout egg eaten by one specimen. 



The relative importance of the difi'erent groups of insects was very ditterent in 

 the different localities. The specimens from the South Fork of the Kaweah River 



