132 



level as the water iu the tank. When the siphon of the outflow tube is empty the 

 water in the tank and jar rises until the level of its surface reaches the bend of the 

 siphon. Then the siphon fills and commences to act and the level of the water sinks 

 gradually until it falls below the short leg of the siphon, when tlie latter empties and the 

 outflow stops. The wide opening at the bottom of the jar is covered with a single 

 thickness of coarse cheese cloth to prevent the escape of the eggs which are introduced 

 into the jars through the narrow opening of the neck above. The rise and fall of the 

 water is only five inches vertically. Eyder says that cod eggs were hatched in this way 

 with a loss of only five per cent. I tried this apparatus with a large number of eggs 

 of the flounder and plaice {Pleuronectes Jlesus and P. platessa). 



J 



m 



Fig. -B. Fig. F. 



Fig. i/. Diagram of a transverse section of an apparatus for hatching buoyant fish eggs, arranged 

 according to the method devised by Captaia Chester, of the United States P'ish Commission. 



Fig. F exhibits the modification of the apparatus adopted in the Laboratory of the Marine Biological 

 Association. 

 y, The glass hatching jar, resting on two bricks in a small tank containing sea- water ; o, the overflow 

 tube ; s, siphon in the overflow tube ; <, an indiarubber tube conducting the inflowing water 

 into the interior of the hatching jar. The dots represent the eggs. 



On February 12 I placed a large number of ova of the flounder, artificially 

 fertilised the same day, in two glass jars of the form described. The jars were 8 inches 

 wide and 17 inches high. The bottom of each jar was covered with silk bolting cloth. 

 Each jar was supported on two bricks resting on the bottom of a shallow aquarium tank 

 made of slate and glass. The water escajied from the tank by a stand pipe into which 

 I fitted a glass siphon so that the level of the water in the tank oscillated between limits 

 about 4 inches apart. The water inside the jar was of coiirse perfectly still : its 

 gradual rising and sinking caused no disturbance in it, in fact I found that the upper 

 part of the water in the jar was scarcely changed or affected by the rise and fall. 

 This will readily be understood. The total height of water iu the jar was about 10 

 inches : at the top of this floated the eggs which formed a layer about \ inch thick. 

 As the level of tire water in the tank sank, the water at the bottom of the jar escaped 

 through the bolting cloth. When the tank was at its lowest level the height of the 

 water within the jar was still 6 inches. When the water rose again all this water 

 was bodily lifted up with the layer of eggs at its surface, while 4 inches of new water 



