MARINE AQUARIUM PROCEDURES AND TECHNIQUES 



59 



cause it to leave at the opposite angle and 

 at the top. If the current is sufficiently 

 strong, and if the form of the tank is ade- 

 quate, then a rapid test performed with a 

 few milliliters of colored water — e.g., by 

 the addition of a few drops of a neutral 

 red or any other nontoxic coloring sub- 

 stance — will enable us to find some eddies, 

 the localization of which can be most use- 

 ful in a breeding tank. 



Of the measures required to keep the 

 water supplied with oxygen in the dif- 

 ferent tanks of a series, we have demon- 

 strated a reduction of that oxygen which 

 progresses from the exhausting device to 

 the lowest tank. To remedy that shortage, 

 we have arranged the tanks, not in a paral- 

 lel series, but in the shape of a fan; this 

 has enabled us to use progressively larger 

 tanks. Accordingly, in the first tier each 

 tank receives the water from one exhaust- 

 ing device. For example, four tanks re- 

 ceive each the water of one exhausting 

 device. The second tier comprises only 

 two tanks, of a larger size. Each tank 

 receives the water of two higher tanks, 

 that is, of two exhausting devices. The 

 third tier comprises only one large tank 

 which receives the water of the two inter- 

 mediate tanks, that is, of four exhausting 

 devices. 



This distribution affords greater safety 

 to the overall installation. As a matter 

 of fact, if one exhausting device stops op- 

 erating for one reason or another, only 

 one single tank will not be fed any longer ; 

 the tank of the second tier that is branched 

 upon it will still receive the water of the 

 exhausting device of the neighboring tank 

 while the tank of the third tier will receive 

 the water of three exhausting devices. 

 Finally, it is possible to double the number 

 of exhausting devices and to feed them by 

 means of two pumps both of which are 

 branched upon all the tanks. 



Connections between tanks — the siphons 



When plastic tanks are used, it is very 

 easy to make an overflow pipe by perforat- 

 ing the wall of the tank and attaching to 

 it a plastic or glass tube. In the latter 

 case, it is preferable to solder it. Any 

 wood or rubber connections should be 

 avoided. The former are easily gnawed 

 by animals — sea urchins, gastropods, etc. — 

 while the latter are frequently attacked by 

 the sea water. It is easy to secure plugs 

 of plasticized polyvinyl chloride; 'they re- 

 main flexible and are not attacked by sea 

 water. (Note well: some polyvinyl prod- 

 ucts are not neutral in sea water.) 



When the breeding tanks are made of 

 glass (our installation is of this type), the 

 perforating of the walls is always a deli- 

 cate operation. The use of siphons makes 

 this operation unnecessary. Over a period 

 of several years, we have used two types of 

 siphons which have been entirely satis- 

 factory : 



1. Permanent siphons. — Generally 

 speaking, the siphons will not run dry as 

 long as the inlets and outlets remain in 

 the water. In the siphon shown in figure 

 3, the inlet A is maintained constantly 

 within the water by means of the mounted 

 swanneck C which connects directly with 

 the outlet of the siphon. An inlet for air 

 has been created at E to simplify priming 

 the siphon between B and D . On the other 

 hand, that air inlet makes it possible to 

 prime the first siphon by aspiration 

 through that opening with the aid of a 

 flexible tube. The water arrives through 

 A and F, filling the tube. When the 

 flexible tube is withdrawn, the part CD 

 will dry up by itself, and the apparatus 

 will be in operating condition. When the 

 end F does not reach into the water of the 

 lower tank, it is necessary to plug it dur- 

 ing the priming operation. 



We have equipped the inlet of certain 

 siphons with a sort of funnel, the opening 

 of which is covered by a nylon screen that 



