44 Papers from the Department of Marine Biology. 



ature) between May 18 and 30, and more narrowly between May 20 

 and 25. There are exceptions, but on the whole the breeding season 

 is seemingly concentrated within narrow limits and the period of 

 incubation has its approximate beginning within these 10 days. 



The information at hand is even less definite as to the time of cessa- 

 tion of the care of the male parent. The mortaUty of eggs hatched 

 in aquaria is very great, and with young eggs all may be expected 

 to die. In MacDonald hatching-jars, which have a bowl-shaped 

 base with water admitted at the center of the base, the eggs are 

 lifted by the incoming current, and all sediment, mold spores, and 

 bacteria are carried off by the overflow at the top. It was thought, 

 moreover, that the use of these jars would conserve the life of the 

 eggs in another way. When the yolk-circulation has extended its 

 plexus of blood vessels over the ventral side of the eggs, if these are 

 allowed to stand in one position, shortly a congestion of these ven- 

 tral blood vessels is noticed and this is quickly followed by the death 

 of the eggs. In the mouth of the father the constant current of water 

 due to the respiratory actions keeps the eggs free from sediment, 

 bacteria, etc., and presumably keeps them in motion, thus preventing 

 the congestion referred to. It was hoped that in the MacDonald 

 jars the uplifting action of the water would keep the eggs "dancing" 

 and prevent this trouble. To some extent this hope was realized. 

 But even in the best-handled jars, although there was little conges- 

 tion, the eggs turned white and died by dozens and scores, and in no 

 case was it ever found possible to carry any but late eggs on to hatching. 

 Eggs brought in with young nearly ready to burst the shell could 

 with great care be carried through this dangerous period, and if car- 

 ried 3 or 4 days further, little trouble was had in bringing them on to 

 maturity. It is interesting to note, however, that there was never 

 any evidence of death due to bacteria, as has been found in the eggs 

 of the toadfish, Batrachus tau. For the details of the difficulties met 

 with in the effort to hatch eggs, the reader is referred to the author's 

 1916 paper. 



However, although no definite time can be set for the liberation 

 of the young from the mouth of the father, this can at least be approx- 

 imated as the time when the body-walls have closed up, as shown 

 in figure 12, plate iv, and the young fish set out to find their own 

 living. In 1907, on July 29, a little catfish was killed and hardened in 

 formalin for 15 hours, at the end of which it measured 81 mm. long 

 oA'er all, while the yolk-slit measured 16.5 mm. long by 2 mm. wide. 

 On August 1, when I left Beaufort, the other fish of this lot in a 

 large aquarium were relatively as active as the adults in Newport 

 River: their babyhood was over. 



In 1908 two sets of larvae were under observation at the close of 

 July. The older ones had the yolk covered over and the slit reduced 



