14 
In discussing the results obtained we must first point out some of 
the most apparent errors in the experiment. The experiment was 
performed during the hours 8 a. m. to 5 p. m. and therefore the 
night life of the fish is not recorded. It is probable that many more 
larve would have been eaten had they been given to the fish very 
early in the morning. Others would have been eaten as late as 7:00 
p. m. and probably some would have been eaten in the darkness of 
the night. Again since the killifish depends almost entirely on its 
visual sense in locating the mosquito larve and since the specimens 
used could get no new lot of 10 larve until all but one larva, dead 
or alive, was eaten, many minutes of idleness and probably hunger 
were passed over. Lastly, the food was entirely of mosquito larvee. 
Briefly summarizing the results of the experiments we may say: 
that first, the highest record of daylight consumption for the period 
of four days was 123 larve; other fish ate more but did not survive: 
second, the highest record of daylight consumption of larvee for three 
of the four days was 1385, and then the fish died; third, the total 
number of larve consumed during four days and three nights by 
four of the fish which survived was 600; the total number consumed 
by these same fish for the daylight hours only was 451; fourth, the 
greatest number of larve eaten in one day was 70, but two of the 
fish made this record and one of them ate but 20 larve the following 
day and then died; fifth, the two largest fish were the only ones 
which survived the whole period of four days and seemed ready for 
more food at the end of the experiment. The average consumption 
of the two survivors per day for the period of four days was 35 
larvee be day each. The average daylight consumption for tha 
time was 27.25 larvee each. 
Tt is obvious that experiments performed on such a small number 
of fish for such a short period of time prove little. They do show, 
however, that the capacity of medium-sized killifishes of the species 
Fundulus heteroclitus is enough to make it a formidable enemy of 
the mosquito larvee. 
It is interesting to note that Fundulus heteroclitus not only will 
eat larve, pups, and the shells of the same, but that it will eat 
these over and over again, when hard pressed by hunger. 
In order to test the activity of Fundulus heteroclitus as an enemy 
of the water tiger, larva of Dytiscus, I performed numerous experi- 
ments at Woods Hole, Mass. .. using medium-sized killifish and rather 
large water tigers. In thé presence of mosquito larve the water 
tigers were not eaten, but after the mosquito larve had been cap- 
tured the water tigers were soon devoured. 
‘As Dr. G. H. Parker (7, 8) has shown the importance of the 
visual sense as well as the olfactory sense in the teleosts, particularly 
Fundulus heteroclitus, experiments on the attractiveness of larval 
shells to Fundulus were hardly necessary. It seemed wise, however. 
to make a few such experiments. 
