Marine Animal and Vegetable Life. 159 



On keeping Aquarium water in darkness for three or more weeks, 

 it undergoes great purification. Thus in one case after twenty-five 

 days there remained only 2 per cent, of the free and 48 per cent, of 

 the albuminoid ammonia originally present. There was no further 

 purification on keeping an additional twenty-nine days. This purifi- 

 cation is due to bacterial action, as it is greatly delayed if the water 

 be previously heated to 100, and is stopped by adding corrosive 

 sublimate to the water. The amounts of free and albuminoid 

 ammonia present after this purification are in some cases no greater 

 than those found in pure open sea water. The physiological purity is 

 not so great, however, as larvae grown in open sea water are increased 

 in size by 16'0 per cent., and those grown in water kept in darkness 

 by, on an average, 7'5 per cent. Also with this latter water the 

 percentage of ova reaching the full larval stage is slightly smaller 

 than the normal. Water kept in darkness contains considerably 

 fewer germs than normal water, and the longer the water is kept the 

 smaller is the number. 



A very considerable purification of the Aquarium water is effected by 

 the layer of bacterial slirne coating the inside of the pipes which con- 

 duct the water from the reservoirs to the rooms. Thus water drawn 

 off at the rate of a litre in ten seconds contained 26 per cent, less 

 free, and 25 per cent, less albunlinoid, ammonia than the reservoir 

 water, and that drawn off at a litre in two hours and forty minutes 

 respectively 82 per cent, and 16 per cent. less. Larva3 grown in such 

 water were, moreover, some 7'8 per cent, larger than the normal. 



On filtering water through asbestos, the albuminoid ammonia is 

 somewhat increased, but larvae grown in the filtered water are some 

 12*6 per cent, larger than the normal. 



Larvae grown in water previously heated to 50, 76, or 100 are 

 increased in size by some 6'2 per cent., and the proportion of ova 

 reaching the larval stage is also increased. This is probably due to 

 the removal of bacteria and other forms of life from the water, 

 though after the first day or two the water contains more bacteria 

 than unheated water. 



Ordinary tank water gave about 1500 colonies per c.c. after 

 twenty-four hours' incubation at 25, and about 3000 after forty- 

 eight hours. Filtration through sand at first diminished the number 

 of germs, but when this became impregnated with vegetable growth 

 they were largely increased. Through sand kept in darkness, very 

 slow filtration increases the number of germs, but a fairly fast one 

 diminishes them. The addition of algae to the water increases the 

 number of germs. 



Larvae grown in water previously fouled by fish, crabs, molluscs, 



and Holothurians are increased in size on an average by 4'1 per cent., 



but with water fouled by sea-anemones and medusae are slightly 



