180 The Philippine Journal of Science i9i7 



an unconfined space they could almost always escape. The male 

 took no part in this canibalistic feast. 



The female was then removed to a separate jar, and the male 

 was left alone with the young for twenty-four hours without 

 food. He showed absolutely no disposition to attack or eat the 

 young, although he must have been very hungry. 



In old females the young shoot out without difficulty, and I 

 have seen them give birth to five young in as many seconds. 

 The first instinct of the young fish is to get away from the vicinity 

 of its parents, but after swimming about for a few minutes, it 

 settles upon the bottom or upon the leaves of the pond weeds 

 to rest. 



COPULATION AND PERIOD OF GESTATION 



The following day (December 9, 1915) this male was again 

 placed in the jar with this female. It was at once apparent 

 that the relation between the two had entirely changed, for the 

 female at once attacked the male and bit him viciously and 

 chased him about the jar. The male very evidently was afraid 

 of her. However, this did not prevent the male from making 

 repeated successful attempts to copulate with the female. These 

 attempts were always made by stealth and without the consent 

 of the female. 



The exact method of copulation was as follows: The male 

 would get behind and a little below the female ; then if she was 

 not watching he would suddenly dart forward, at the same time 

 turning forward the large modified anal fin, which functions 

 as an intromittent organ, and would attempt to insert this organ 

 in the cloaca of the female, who would at once turn and fiercely 

 attack her would-be mate. These exchanges continued irreg- 

 ularly for about three days and were gradually given over by the 

 male. 



Eight weeks later (February 3, 1916) this female showed 

 decided signs of pregnancy, and three weeks later (February 25, 

 1916) the male was seen to be making passes at the female and 

 biting at her lower abdomen. He seemed to have lost all fear 

 of her, while apparently she sought only to escape. This change 

 of attitude of the sexes seems to be an unfailing sign that the 

 spawning time has arrived. In one hour this female gave birth 

 to 48 young. The methods followed and the actions of the adult 

 fish were the same as previously described for this pair of 

 fish on December 8, 1915. This second spawning establishes the 

 fact that the period of gestation for this species is not more than 

 seventy-nine days. During this period this pair of fish ate 5,041 

 mosquito larvse by actual count. 



