MOSQUITO-DESTROYING FISH 407 



other than the mosquito. The black-banded snnfish, Mesogonistius chcetodon, 

 would also be desirable for this purpose, if they were not so difficult to obtain in 

 large numbers. One or both species of Enneacanthus can be found wherever 

 there are aquatic plants. The above-mentioned five species in combination seem 

 to be the most suitable for pond protection of all those which are known to 

 thrive in still water, and which in any degree possess the desired qualities. As 

 has been stated, the killifishes would probably be found to be undesirable. In 

 their natural habitat, the tidal streams and great expanses of small marsh, their 

 efficiency is unquestioned. 



" There are many places at the seashore where there are swales or hollows 

 filled with grasses and bushes, which in periods of rainfall become breeding 

 places for the mosquito, especially of Anopheles. If these places are stocked with 

 fish, the result is that when they dry up the fish perish, and the operation must 

 be repeated after each filling. 



" The writer has suggested digging holes about four feet square down through 

 the turf into the sand stratum in the deepest part. Two feet is usually sufficient 

 to secure a constant water supply where the fish can exist until the hollow is again 

 rain-filled. Cyprinodon and Lucania would be desirable for such places, and 

 they are to be found everywhere in the ditches and tide pools on the flats. 



" To add variety to the treatment of the subject, it might not be amiss to 

 suggest that there is a fish, Andblops, inhabiting the fresh waters of South 

 America, which seems to be specially adapted to this purpose. To quote : ' These 

 small fishes swim at the surface of the water, feeding on insects, the eye being 

 divided by a horizontal partition into a lower portion for water use, and a portion 

 for seeing in the air.' " 



Acting largely upon Mr. Seal's advice. Doctor John B. Smith, the State 

 Entomologist of New Jersey, with Mr. Seal's help, in November, 1905, brought 

 Gamhusia affinis and Heterandria formosa from North Carolina into New 

 Jersey, which were distributed as follows : 8000 in spring and natural drainage 

 rivulets flowing into the ice pond at Westville, N. J., 600 in a landlocked pond 

 near Delanco, 600 in a mill pond between Merchantville and Evesboro, 600 in 

 landlocked waters near Delair, and 400 in ponds of the Aquarium Supply Co., 

 at Delair. In Doctor Smith's report for the year 1906, it was stated that the 

 experiment was in large part a failure. Whether it was due to the destruc- 

 tion of the introduced fish by black bass, pike, yellow perch and sunfish, or 

 whether because of other enemies, or because of their dislike to their changed 

 conditions they found their way during the spring rains to rivulets flowing to 

 the Delaware Eiver, or whether they escaped in other ways, could not be told. 

 In hie report for 1907, however, Doctor Smith states that the Gamhusia was 

 found in large numbers in Teal's Branch of Pond Creek, a small tributary' of 

 Delaware Bay at Higbee's Beach, by Mr. Henry W. Fowler, of the Academy of 

 Natural Sciences, Philadelphia, and Messrs. H. Walker Hand and 0. H. Brown. 

 These gentlemen found it also very abundant in New England Creek, another 

 tributary of Delaware Bay just north. Doctor Smith states that Mr. Seal was 

 inclined to claim that this finding was the result of his work in 1905, but that 

 Mr. Fowler doubted this conclusion since the points where found were 90 miles 

 distant from points of introduction. 



Although the opinions of Mr. W. P. Seal, here quoted, are authoritative, it 

 may not be out of place at this point to quote the interesting account by P. E. 



