ARTIFICIAL PROPAGATION" OF SHaID 483 



With a proper quantity of semibuo3^ant eggs in the jar and the 

 water turned on and regulated, the movement of the current estab- 

 lishes a regular boiling motion in the mass of eggs, wliich brings each 

 in succession to the surface. Tliis motion may be regulated without 

 altering the quantity of water. By loosening the upper binding 

 screw of the central stuffing box and pushing the feed tube down 

 until it almost comes in contact with the bottom of the jai- the motion 

 of the eggs is increased. If the jar is working properly, the dead eggs 

 when brought to the surface remain on top, forming a distinct layer, 

 and by pushing the outlet tube a suitable distance they are carried 

 off by the escaping current. 



When the water is turned on for the first time, the jar should ])e 

 watched closely until a legular motion has been established. When 

 eggs have stood 15 or 20 minutes in the jar before the water is turned 

 on, they do not readily yield to the boiling motion but tend to rise 

 in a solid mass to the top of the jar. By quickly starting and stop- 

 ping the current the mass is readily disintegrated. Tiie degree or in- 

 tensity of motion of the ego;s varies not only with their age and condi- 

 tion, but also with the condition of the water. If the water is mudd}'", 

 the motion should be rapid enough to prevent mud settling either on 

 the eggs or in the bottom of the jar. Ordinarily the best motion is 

 that which readily brings the dead eggs to the surface. After the 

 hatching has progressed far enough to dispose of a portion of the eggs 

 there is less resistance to the current, and it should be reduced by 

 shutting off part of the supply or by slightly lifting the central tube. 

 If the motion is not reduced from time to time as the hatching pro- 

 gresses, shells will be carried over into the receiving tank with the lish 

 and, being very light, will be drawn against the outlet screen, causing 

 an overflow. The motion should be so gentle at the time of the great- 

 est hatching as barely to induce the fish to swim out of the jar and 

 leave their cast-off shells behind. 



If the connection of the jar must be broken, it is essential that the 

 rubber feed tube does not drop down and siphon the eggs from the jar. 



In reconnecting, the air may be expelled with the metal top screwed 

 down in position. To effect this, draw both glass tubes up to the top 

 of the jar and turn on a full head of water, when the air will be forced 

 out in bubbles above the eggs, the bub])les escaping through the out- 

 let tube. The central tube is now restored to its former position. 

 The automatic action permits entire separation of bad from good eggs, 

 though some days may be reciuired to accomplish the full result. 

 The dead become lighter from gases arising from decomposition. 

 A net small enough to enter the mouth of the jar easily and fixed to 

 a handle several inches longer than the jar is convenient for removing 

 particles of foreign matter. 



At the approach of the hatching period a careful watch must be 

 kept on the screens guarding the overflow in the fry tanks to pre- 

 vent clogging through an accumulation of eggshells carried over from 

 the hatching jars. The application of a stiff brush to the inner sur- 

 face of the screen is the best way to overcome this difliculty. Usu- 

 ally a surplus quantity of the shells will remain in the jars, forming 

 a cloudlike layer above the moving mass of eggs. This may bo elim- 

 inated from the McDonald jar by manipulating the outlet tube, as 

 already indicated, or, in the case of the open-top jars, by the siphon 

 described below. 



