656 NEW JERSEY AGRICULTUEAL COLLEGE 



that. tJicy will gladlv nndcrtaki' it if sd autliorizi'(_l, and that with 

 tlieir extensivo eqiiipiiu'iit and facilities the experiment can be 

 made on a scale of such maiiuitude that success will be assured. 



March, or not lat<>r than Ai)ril, will be found to be the best sea- 

 son for shiiviunu' thciu hy express. If acconi}ianied bv a lish eom- 

 niissiou niesseuiicr. ^lay would not be tno late. At all events the 

 sprinii' mouths are ihe ]n"oper ones for this work. The earlier the 

 better. Lheii with lh(> entire s])rin<;- and sunnner for breeding, the 

 multiplication will bc^ manifold. 



Gmnbusia affiius and Heiemndrui. formosa< arc the smallest 

 Oj'prinodonts found on the Atlantic coast. They are viviparous, 

 surface-feeding", la.rva3-eating, minnows, know as "top-miiuiows..'^ 

 A study of the breeding habits of a pair of Hetemndrla during the 

 past summer showed the extrusion of young at intervals averaging 

 a week from dune loth to Septemlx^r 10th. some fourteen deliv- 

 eries in all. These varied in the number of young extruded, but 

 a fair average would bo ten, although as high as eighteen were 

 found on on(> occasion. Some were undoubtedly eaten when other 

 food was scarce, though in the cases observed it might have been 

 a parallel to what has been observed of other species where the 

 yoimg appeared to be devoured because they were deficient in 

 vitality. Taking it, however, at the average of ten, there w^ould 

 Ik' an iucreasi' from \\w (Uie pair of one hundred and forty. Xor is 

 this all. A pair horn in June brought forth young in August and 

 another pair did tlu^ same in September, thus ]>roducing a second 

 generation within four months. It will thus be seen that they ai'e 

 wondi'ously ]>rolitic. Although they do not have the enormous 

 fccni'dity of most tishes. the young are protected during the e^ 

 stage ai«l are born able to take up the burdens of existence with 

 a much grc^atcr chance of escape from tlieir enemies as they are 

 fully dev(doped and very active at birth. 



It can bo safely assumed that the breeding habits of Gamhusla 

 are identical with those of Ileterandna. 



It is an established fact that these two s^x^eies are largely feeders 

 upon the Inrvip of mosquitoes, and that because of their small size 

 and habit of swimming and feeding at the surface, they can pene- 

 trate among and over aquatic plants as no other species can or do, 

 and for this reason are especially adapted for the destruction of the 

 larvie of Anopheles. This phase of the question has, however, been 

 quite fully explained and should Ih? fully understood. 



