MINNOWS GAMBUSIA AFFINIS AND CYPRINODON VARIEGATUS. 13 



period another attack is made, and this is kept up until the victim 

 is exhausted or disease attacks the wounds. In several instances 

 it was noticed that a number of individuals made a concerted attack 

 upon one common victim. Where the prey is large and can not bo 

 devoured whole, the flesh is ripped from the bones with the sharp 

 teeth and eaten a bit at a time. Cyprinodon does not Hmit its attacks 

 to fishes of other species. When a number of them are placed in 

 an aquarium, fighting soon ensues among their own kind and can- 

 nibalism prevails. 



It is apparently a voracious feeder, with a varied diet. In nature 

 it appears to subsist largely upon vegetable matter. The stomachs 

 that were examined were found to be distended with plant stems, 

 algje, and mud. The nature of the digestive tract, which is much 

 convoluted and equal to about two and one-third times the length of 

 the fish, indicates that plants form the principal natural food. 



In 1914 it was noticed that this fish spawned throughout the 

 summer, so that ripe females could be obtained at nearly any time 

 from April, when the observations were begmi, until October. It was 

 also found that there were several sizes of eggs present in the ovary 

 at one time. These facts suggested that this fish produced more than 

 one set of eggs durmg a single season. In order to obtain more defi- 

 nite information in regard to this matter, the following experiments 

 were undertaken: A rectangular box was constructed with four legs 

 and with a hole in the bottom near one end. Beneath the hole there 

 was tacked a piece of wire netting, the meshes of which were too 

 small to permit the escape of the fish to be used in the experiment, but 

 large enough to allow the eggs to pass through, should any be pro- 

 duced. This box was placed in a compartment of a hatching table 

 provided with an overflow and connected with drain pipes. Under- 

 neath the hole in the bottom of the box a small basket of wire gauze 

 was placed. The opposite end of the box was somewhat elevated. A 

 small stream of salt water was allowed to flow in at the elevated end, 

 thus creating a current directed toward the opening in the bottom at 

 the opposite end. The purpose of the current was to carry the eggs 

 through the screened opening and cause them to be deposited in the 

 small gauze basket. On April 10 a large female was placed in the 

 box, where she lived until September 20. Eggs were produced on the 

 following dates: April 28, May 28, June 14, daily from June 24 to 

 July 3; July 16, 17, 19, 22, and daily from July 24 to 31; August 9, 

 11, 13, and 16. The first three sets consisted of from 18 to 24 eggs 

 each and the remainder of only from 2 to 6 each. When the female 

 died on September 20 she was completely ' 'spawned out." It is, how- 

 ever, probable that aU of the eggs that were produced did not reach the 

 retaining basket, as dissections indicate that the early sets at least are 

 usually much larger. It is possible that the parent may have eaten 



