382 Proceedings of the Royal Physiccd Society. 



renewed by forcing air into the water, and by the second a 

 continuous stream of water was kept flowing through the 

 tanks. The first method, although quite satisfactory on a 

 small scale, decreases in its applicability just as the size of 

 the tank increases. The second, although perhaps more 

 natural than any other method, is very expensive, and not 

 applicable for an inland station. It is necessary to pump 

 the water direct from the sea, and to return it again after 

 having circulated through the tanks. This method also has 

 another disadvantage. It is found by experience that water 

 kept properly aerated improves instead of deteriorating with 

 keeping. I'orms which wdll not live in water newly intro- 

 duced into a tank may often be kept there with ease six 

 months afterwards. Again, that crystal appearance so 

 characteristic of pure sea water is never to be obtained from 

 water freshly pumped from the shore. Practically a com- 

 bination of the two methods is most commonly in use. A 

 small jet of water is forced into the tank with such a 

 pressure as to carry air along with it. 



I propose, then, to give an account of the methods and 

 apparatus by which experience has shown that this can best 

 be ahected with due consideration of expense. When some 

 four or five years ago I had my aquarium fitted up in 

 Huddersfield, I was supplied with vulcanite pump, pipes, 

 and jets, by Messrs Leete, Edwards, & Norman, similar to 

 those supplied to the large public aquaria. I took the pre- 

 caution, however, to have every jet supplied with a stopcock, 

 so that I could vary the amount of water issuing from each 

 jet at will. I have never regretted this precaution, and can- 

 not too strongly recommend its general adoption in all marine 

 laboratories. The extra expense is amply repaid by the 

 extra convenience obtained. In discussing the various altera- 

 tions in the system of aeration I have made from time to 

 time, the subject seems naturally to fall under two heads, 

 viz., the pumping apparatus used, and the various forms of 

 jets experimented with. The economical and efficient result 

 obtained is of course dependent on the combined action of 

 the two sets of appliances, but it will be well to consider 

 them separately. 



