The Relations between Marine Animai and Vegetable Life. 353 



to the layev of slime or zooglcea which gradually forms od the suv- 

 face of the sand, as a deposit from the filteving- water. Thus it is 

 known that a fresh sand hlter takes several days to attaiu its full 

 filtering effìciency. 



In order to try the effect of such filtration on the Aquarium 

 tank water, a stream of water was allowed to run slowly, at a rate 

 of about two litres }3er hour, through a layer of moderately eoarse 

 sand 14 cm. in depth, and of about 100 sq. cui. superficial area. The 

 sand used in this experiment had been taken from one of the tanks 

 in the Aquarium. The follo wing was the change in the composition 



of the water. 



Free NH3 Organic NH3 



Before filtration .268 mgm. .165 mgm. 



After filtration .017 .136 



That is to say, by simple filtration through sand, no less than 

 94^ of the free ammonia present in the water, and 18^ of the 

 organic ammonia had been removed! It seemed impossible that such 

 purification could be due merely to mechauical filtration. That this 

 was obviously not the case was proved by a microscopical examination 

 of the sand. It was theu fouud that every grain of sand was covered 

 with a thiu layer of structureless particles of alga, and bere and 

 there with diatoms. The removal of the ammonia must bave been 

 due to this vegetable agency. That this was so was confirmed by 

 another experiment, made with some of the sanie sand, which had 

 nowever been left stagnating in a jar of water for ten days, and 

 which probably on this account did not act so efficiently. Thus ou 

 filtering water through a layer of this sand, 11 cm. in depth, 79^ 

 of the free, and 37^^ of the organic ammonia, were removed. Some 

 of this sand was then heated to 70° C. with sea-water, by which 

 means ali the vegetable life present was killed. It was then washed 

 several times, to remove as much of the dead vegetable debris as 

 possible. On filtering water through a layer of this water sand, 7 cm. 

 in depth, so far from the ammonia present in the water being 

 dimiuished, it was found that the free ammonia had increased 18^, 

 and the organic ammonia 272^;, or nearly four fold. 



With a view to making a more accurate estimate of the conditions 

 of water purification by sand, an apparatus consisting of a large 

 inverted tubulated bell-jar was set up. Over the tubulure of the 

 bell-jar was placed a small porcelaiu sieve, and on the top of this 

 a small inverted glass dish. The bell-jar was theu fiUed to a depth 



