SUNIER: The Laboratory for Marine Investigations. 143 



Finally each of these two pipes (at T and T in the lower half of 

 the Diagram) discharges into a reinforced concrete duct which connects 

 the centres of the East and West wall of a filter tank. 



The floor of this duct on the inside is about level with the top surface 

 of the quartz sand layer. 



The side walls of these ducts are pierced by numerous holes, at 

 different levels and along the whole length of the ducts. 



By means of this, and independently of the amount delivered, the water 

 is always spread equally over the filtering sand surface in any case in the 

 direction from West to East. 



The outlet of water from the bottom of the filter tanks to the bottom 

 of the mixing reservoir takes place in exactly the same way as is the 

 case with the tanks in the aquarium chamber. 



Only the inlet and outlet arrangements of the filter tanks and reservoirs, 

 which also do not act as syphons but as overflows, are not composed 

 of earthenware pipes, but of plastered brickwork. 



A comparison of the Diagram with Plan No. 4 shews sufficiently 

 clearly how the sea water flows from the surface of the mixing reser- 

 voir (2, lower half of Diagram) to the bottom of the first large Souther- 

 ly reservoir (3, lower half of Diagram), and from the surface at the other 

 end of that reservoir again to the bottom of the second large Northerly 

 reservoir (4, lower half of Diagram), only finally to be pumped up again 

 from the other end of the last-mentioned reservoir (right-hand lower cor- 

 ner of the Diagram, at D). 



The distilled water reservoir and the water distilling apparatus itself 

 are indicated on the Diagram by the numbers 4' and 5'. 



From the measurements given above it appears that the combined 

 capacity of the reservoirs and filter tanks is not less than fully eight times 

 as large as that of the tanks in the aquarium chamber. Even if we neglect 

 the capacity of the filter tanks, which are for the greater part taken up by 

 stones and sand, the combined capacity of the mixing reservoir and the 

 large reservoirs is still seven times larger than that of the tanks in the 

 aquarium chamber. 



In the literature available to me here, I have been able to find nothing 

 about the theory of the purifying of the water in the cellar of an aquarium, 

 system Lloyd, in which the same water is always circulating. 



It appears, however, not unlikely that the filters operate also as 

 oxidation beds and that the nitrates and nitrites formed on these oxidation 

 beds are denitrified by bacteria in the darkness of the large reservoirs. 



The reinforced concrete portion of the aquarium was built by the 

 Hollandsche Beton Maatschappij of Weltevreden. 



The pumping sets and the earthenware pipes, which latter were manufac- 

 tured by the N. V. Nederlandsche Gresbuizen Industrie at Deventer, were 

 delivered by the Amsterdamsch Kantoor voor Indische Zaken in Batavia. 



