December 27, 1901.] 



SCIENCE. 



1017 



From the above brief description it will 

 readily appear that the Sauropod manus in 

 Brontosaurus at least was like the pes, ar- 

 ranged on the entaxomic plan, and not mesax- 

 omic as Professor Osborn was led to believe 

 from a study of the abundant but isolated ma- 

 terial in the collections of the American Mu- 

 seum of Natural History. 



This foot and limb will be more fully de- 

 scribed and adequately illustrated in the com- 

 ing number of the Annals of this Museum. 



J. B. Hatcher. 

 Cabnegie Museum," 

 December 2, 1901. 



GOLDFISH AS DESTROYERS OF MOSQUITO 

 LARV^. 



In Professor L. O. Howard's recent excellent 

 volume on 'Mosquitoes,' etc. (p. 161), reference 

 is made to a reported employment of * carp ' as 

 destroyers of mosquito larvae and doubt is ex- 

 pressed as to the facts in the following words : 



" It was stated a number of years ago in In- 

 sect Life, that mosquitoes were at one time very 

 abundant on the Riviera in South Europe, and 

 that one of the English residents found that 

 they breed abundantly in the water tanks, and 

 introduced carp into the tanks for the purpose 

 of destroying the larvae. It is said that this was 

 done with success, but the well-known food- 

 habits of the carp seem to indicate that there is 

 something wrong with the story. If top-min- 

 nows or sticklebacks had been introduced, how- 

 ever, the story would have been perfectly 

 credible, and it points to the practical use of 

 fish under many conditions. " 



An examination of Insect Life (Vol. IV., p. 

 223) and also of Nature (Vol. XLIV., 1891, p. 

 591) — the origiual source of the statement in 

 question — fails to reveal the precise species of 

 ' carp ' here referred to, but if, as seems likely, 

 it was the common goldfish, Carassius auratus, 1 

 happen to be in a position to confirm the gen- 

 eral truth of the story. 



About six years ago at my home in Belmont, 

 near Boston, Massachusetts, I constructed a 

 small artificial pond in which to grow water- 

 lilies and other aquatic plants, and also to breed, 

 if possible, some varieties of goldfish — though 

 the latter object was a secondary consideration. 



The advisability of making this pond had been 

 somewhat questioned on account of its close 

 proximity to my house and the fact that such 

 ponds are likely to become excellent places for 

 the propagation of mosquitoes. Nevertheless, 

 the plan was carried out and the pond was 

 stocked with goldfish taken from natural ponds 

 in the vicinity where they had been living and 

 breeding, to my personal knowledge, for a 

 number of years. 



The aquatic garden has proved a success and 

 the goldfish have meantime thriven and multi- 

 plied. Moreover, no mosquitoes attributable 

 to the pond have appeared and I have been 

 unable to find any larvse in it, although I have 

 searched repeatedly and diligently for them. 

 I have always believed that the absence of 

 mosquito larvse from this pond was due to the 

 presence of the goldfish, and I have so stated 

 in a paper ' On the Drainage, Reclamation and 

 Sanitary Improvement of Certain Marsh Lands 

 in the Vicinity of Boston ' in the Technology 

 Quarterly, XIV., 69 (March, 1901), as follows: 

 "Iq the water [of this pond] are hundreds of 

 goldfish that feed upon the larvse of mosqui- 

 toes and serve to keep this insect pest in check." 



On observing the statement referred to in 

 Professor Howard's book I determined to make 

 careful observations, to settle the point in dis- 

 pute. Within fifty feet of the pond in question 

 stands a large tank which for a long time last 

 summer was filled with rain water. Here I 

 found constantly large numbers of mosquito 

 larvse, of both Culex and Anopheles. Between 

 this tank and the pond runs a cool brook, fed 

 by a spring' Here also I found abundant mos- 

 quito larvse, those of Anopheles being more 

 prevalent than those of Culex. Reflecting upon 

 this fact it seemed more probable than ever 

 that the goldfish were holding the mosquitoes 

 in check in the artificial pond while in the 

 brook the insects were breeding in compara- 

 tive safety. 



To test the correctness of my theories I took 

 from the pond a small goldfish about three 

 inches long and placed it in an aquarium where 

 it could, if it would, feed upon mosquito larvse 

 and still be under careful observation. The re- 

 sult was as I had anticipated. On the first day, 

 owing perhaps to the change of environment. 



