94 Seal — Breeding Habits of Viviparous Fishes. 



minutes to as many hours. The number given off at one time 

 varies also. After the extrusion of such young as have been 

 developed, there is a period of rest of several days during which 

 another lot is developed, a limited number of eggs only being 

 fertilized at each sexual contact. I can not do better than to 

 copy a series of notes taken from my observations. 



1906. July 14 to 16. Heterandria female extruded 16 young. 

 Some appeared on 14th, some on 15th, and some on 16th. Last 

 previous delivery from same female was seven young. Three 

 prior deliveries at intervals not noted, but not greater than the 

 last mentioned one, resulted in the delivery of five, seven, and 

 seven, respectively. It is impossible to say whether those num- 

 bers represent accurately the respective deliveries, as some were 

 born at night and at other times when they were not under 

 observation, and some may have been eaten. 



July 21st (five days later). Found 9 young and on 23d 

 found 2 more. Saw the female eat one of these. This may 

 only occur when other animal food is wholly absent, or in the 

 case of such young as are imperfect — deficient in vitality or 

 deformed in some way. On the other hand there seems to be 

 a universal tendency on the part of the females of all nest build- 

 ing fishes to devour their eggs and young. This may also be 

 characteristic of females of viviparous species. The writer has 

 recorded the action of male sticklebacks in frightening their 

 young by darting at them as though to devour them, causing 

 them to rush to cover among the plants. This is undoubtedly 

 a continuance of the protective habit from the eggs to the 

 young until they are capable of self protection. 



August 1st (8 days interval), found 4 young. August 2d, 

 found 12; August 4th and 5th, 2 more, making 18 in all. 



August lOth-llth (5 days interval), found 4 more. 



August 18th (7 days interval), found 10. 



These fish grow and develop very rapidly if properly fed. 



I had paired two that had been hatched early in May. On 

 July 27th I found one young in the jar but at first supposed it 

 must have been in the jar hiding among the plants when the 

 breeders were introduced, as I could not conceive of two genera- 

 tions of fish in one season, it being so contrary to the habits of 

 fishes in general. But on August 18th I found 6 young about 

 which there could be no doubt. 



