190 BULLETIN OF THE UNITED STATES FISH COMMISSION. 



extent the perfect working of the jar. The glass foot which is shown 

 in the improvised form has been omitted in the form now in use, the 

 jar being supported upon a tripod of three glass lugs; this form of at- 

 tachment being adopted to prevent the distortion of tbe bottom of the 

 jar, which would necessarily result from the attachment of a single foot 

 to it. 



The top of the jar is made with threads to receive a screw cap, and 

 both the bottom and the top surfaces are ground so that the plane of 

 each shall be perpendicular to the axis of the jar and so that when the 

 jar is resting upon its feet its axis shall be perfectly vertical. 



These are all-important considerations to secure the proper working. 

 The top of the jar is closed by a metallic disk perforated with two |-ineh 

 holes — one perfectly central, which admits the tube that introduces the 

 water into the jar; the other equally distant from the central hole and 

 from the edge of the plate. A groove in the inner surface of this metal- 

 lic plate carries a rubber collar, and when the plate is in place the tight- 

 ening of the metallic screw cap shown in the figure seals the opening 

 hermetically. Both the inlet and outlet tubes pass through stuffing- 

 boxes, by which means the tubes can be slid up and down easily and 

 tightened firmly in any desired position. The construction of the jar 

 is such that when the metallic disk is in place the central tube takes 

 the central position necessarily; by loosening the screw cap of the stuff- 

 ing box the central tube can be slid up or down so as to produce just 

 such movement of the eggs as is desired. If the quantity of water 

 entering be small, or the head of water slight, without changing the 

 feed of water we may vary at will che force and velocity with which it 

 enters the jar. By pushing the tube down so as to be almost in con- 

 tact with the bottom of the jar we make a relatively small quantity of 

 water do the work of a larger quantity in producing motion. More- 

 over, as in the season of shad hatching, a full supply of water is needed 

 and not a great deal of motion, this is arranged for by increasing the 

 feed and raising the lower end of the central or supply tube, so that the 

 delivery of the water from it will be under less pressure. This central 

 tube is connected by a rubber pipe with the pet cock which furnishes 

 a supply of water under a constant head. 



The exit tube serves a double purpose — first, as an outlet for the water; 

 and secondly, at our i>leasure to remove the layer of dead eggs from the 

 surface. This is accomplished at stated intervals, say once in twenty- 

 four hours, by loosening the screw of the stuffing box so that the tube 

 will slide readily, pushing it down until the dead eggs nearest to the 

 lower end are seen to begin to pass off. By allowing it to remain in 

 this position a few minutes the layer of dead eggs is swept off entirely. 

 They may be either allowed to pass off in the waste, or better, collected 

 by screens and fed to the fish in the aquaria, thus serving the double 

 purpose of preventing the fouling of the water and furnishing a very 

 appropriate food for many varieties of fish. When the period of hatch- 



