PURIFICATION. 269 



expand their feathery tufts in beauty, and the animals 

 begin to take possession of their holes and corners, 

 and to find themselves at home. But you must lay 

 your account with the loss of some specimens ; some 

 ■will certainly die in the course of the first twenty-four 

 hours, others in the first week. But those which 

 survive the first ten days may be considered as pretty 

 ■well established. 



It is during this period that the grand trial of the 

 experiment usually occurs. There is generally a 

 large amount of animal matter attached to the sea- 

 ■weeds, shells and stones, which are received fi:om the 

 sea, such as minute Annelida, Mollusca and Zoo- 

 phytes : very many of these creatures are already dead, 

 or die immediately ; but being too minute to be 

 detected and removed in detail, they decay, and pre- 

 sently contaminate the water. The first symptom of 

 this is a slight dimming of the crystal translucency, 

 which if unchecked soon increases to a milky white- 

 ness, accompanied by a fetid odour, and terminates in 

 the death of the whole animal collection. 



Purification. — As soon as this begins to be per- 

 ceived, the whole water should be drawn oS'by means 

 of a siphon, without disturbing the sediment, into 

 pans, into which for the present, the plants and ani- 

 mals may be put. The Tank should be wiped out 

 and rinsed, and then the water should be filtered back 

 into it. This is a very simple process : a funnel (if 

 of glass, earthenware, or gutta-percha, the better) is 

 placed over the tank, with a bit of sponge pushed 

 lightly into the top of the tube, so as to allow the 

 water to run through in a narrow thread-like stream. 



