[3] 



WORK OF THE STEAMEE FISH-HAWK. 



watching the circulation and found that the surface water remained 

 practically unchanged, the circulation taking i)lace in the lower portion 

 of the cones. 



To obviate this defect the auxiliary feed-pipes were removed from the 

 base of the cones and laid in the space outside of the perforated plates 

 in such a manner as to give the surface water a slightly circular motion, 

 and a feed-pipe was attached to the siphon tank, thus giving a feed at 

 both top and bottom. 



We labored under the disadvantage of having no eggs with which to 

 experiment, but we knew their specific gravity and utilized such sub- 

 stitutes as small pieces of beeswax about the size of cod eggs, bread 

 dust, &c., which served at least to demonstrate what effect the various 

 movements would exert on minute floating bodies. 



The experiment of admitting feed water at both ends of the cones re- 

 sulted in a complete change of water, but was not otherwise satisfactory, 

 as the circular motion imparted by the surface feed caused a movement 

 of the particles representing eggs towards the center, and, the bottom 

 feed being converted into a discharge while the sijihon was in operation, 

 a miniature whirlpool developed sufficient strength to draw the eggs to 

 the bottom and thence through the discharge pipe to the tank where 

 they were taken up by the siphon and carried to the general discharge. 



Numerous experiments were tried with varying success until, finally, 

 the following arrangement was adopted as most nearly producing the 

 required movement : 



FiCr. 3. 



