12 



JAMES W. ATZ 



tion and of maintenance may preclude this 

 ideal arrans^ement, especially if large 

 tanks are involved. When the units have 

 been small enough (less than 1,000 gal- 

 lons) that they could be provided adequate 

 circulations by means of air lifts, indi- 

 vidual sea-water circulatory systems have 

 been set up in public aquariums, for ex- 

 ample, at "Wuppertal (Wiedemann, 1943) 

 and Bern (Hediger, 1944). Ingenious 

 modifications of filters and aquariums 

 making them more suitable for individual 

 circulations have been described by Wiede- 

 mann, by Chin (1959), and by Kelley and 

 Moreno (1961). 



In some sea-water systems, there is no 

 filter; instead the water is allowed to pass 

 slowly through a large settling tank or 

 reservoir. Plerald et al. (1962) have com- 

 pared this type with the one using filtra- 

 tion, and they indicate that the latter car- 

 ries a greater proportion of animals, viz., 

 about 50 gallons of sea water per pound 

 of living animals as compared with about 

 450 gallons for a system depending on 

 sedimentation alone. Wilson (1952,1960) 

 has described a typical filterless system. 

 In his later paper, he stresses the im- 

 portance of having a reservoir that is cor- 

 rectly designed, and he provides the de- 



tails of a successful one now in use at the 

 Plymouth Aquarium. One advantage of 

 sedimentation over filtration is that filter- 

 feeding invertebrates do better in un- 

 filterecl water (Wilson, 1952). 



Filtering tests that were made on beds 

 consisting of uniform-sized silica sand by 

 William E. Kelley of the Cleveland 

 Aquarium (unpublished report) and of a 

 graded series of quartz or quartzlike fil- 

 trants by Herald et al. (1962) showed that 

 practically all of the filtering was accom- 

 plished in the top few inches of the bed. 

 Conventional deep filters would therefore 

 appear not to be necessary in sea-water 

 systems. Saeki (1958), however, has de- 

 termined the amount of sand that will pro- 

 vide space for the bacteria necessary to 

 oxidize the waste products of a fish, and 

 he states that as a minimum the filtrant 

 should weigh 30 times as much as the fish. 

 Evidently volume as well as surface 

 should be considered in calculating filter 

 requirements. An example of a shallow- 

 bed filter that uses a uniformly sized 

 filtrant is shown in figure 3. 



Hinton (1958) has discussed the prob- 

 lem of an optimum rate of flow in a sea- 

 water circulation and concludes that it is 

 difficult to have too much sea water. The 



BAFFLE HANGER 



Figure 3. — Cross section of shallow-bed filter seen in figure 2. The spray header acts as a manifold 

 aerator, and the baffle can be reversed to direct the water into the opposite filter chamber. 

 Backwashing water is introduced into the collecting drain line through its elevated open end 

 (seen in figure 2). 



