10 U. S. BUREAU OF FISHERIES. 



condensed by grouping the stomachs in lots and recording the con- 

 tents under comprehensive taxonomic groups. The percentages 

 given were determined volumetrically by means of a short section 

 of a graduated burette tube or, when the amounts were very small, on 

 u plankton counting slide. However, owing to the usually frag- 

 mented condition of the remains and to the difficulty of completely 

 assorting them and of separating the mucous and other foreign mat- 

 ter, they have no accurate quantitative value. In column 6 of the 

 tables the first numeral indicates the niunber of stomachs in which 

 mosquitoes were detected ; the second, the total number of mosquitoes 

 counted. When all of these data are brought together and com- 

 pared, they present a significant and in some cases a conclusive body 

 of direct and circumstantial evidence. 



During the course of the investigation ol)servations were made 

 upon 29 species of fishes belonging to 9 families. Some of these have 

 only an indirect or casual relation to the mosquito problem, others 

 a very direct and important relation. In the case of some of the 

 larger species the young only were studied, the habits of the adults 

 being such as to remove them from the category of mosquito destroy- 

 ers. Most definite conclusions were reached concerning the 9 species 

 which receive special consideration in the following pages, viz, roach 

 or golden shiner, goldfish, mud minnow, common killifish, translu- 

 cent killifish, top minnow, blue-spotted sunfish, long-eared sunfish, 

 and common sunfish. 



ROACH OR GOLDEN SHINER (Abramis crysoleucas). 



This species, with its subspecies hosci, is distributed over practically 

 the entire United States east of the Rocky Mountains and is nearly 

 cA^erywhere abundant. In the region of these investigations, cover- 

 ing roughly a considerable part of the Delaware River and a portion 

 of the lower Hudson River drainages, it is ubiquitous and found in 

 waters of nearly all kinds — in the larger creeks and rivers, both tidal 

 and upland, in the pools of small rivulets, in natural lakes and ponds 

 of all sizes, in reservoirs, dams, water-filled quarry holes and clay 

 pits, and in ornamental ponds. Everywhere it is one of the most, if 

 not the most, abundant of fresh-water fishes. The black-striped 

 young in small schools of scores or hundreds, often associated with 

 various minnows and other small fishes, parade and explore the 

 shallows on all kinds of bottom except where densely grown with 

 plants. The adults are found in deeper waters, in the channels, about 

 the mouths of creeks, at dams, at the openings of sluices, and along the 

 borders of dense growths of vegetation where they breed. Neither 

 the adults nor the young, although they thrive in weedy ponds and 

 even appear to prefer the vicinity of vegetation and though always 

 active, are adepts at penetrating dense vegetation or worlring into the 

 little pools and pockets of irregular shore lines. The young are very 

 timid and are always ready to turn and run upon the slightest alarm. 

 Possibly it is for this reason that although frequenting the shallows 

 they seldom glean to the actual shore line. The roach is exceedingly 

 prolific and propagates successfully in a variety of waters. 



Several accounts of the food of the roach have been published. The 

 results reported from localities in New York, Illinois, Michigan, and 



