56 



YVES B. PLESSIS 



undergoes a very strong alkaline reaction 

 which makes the use of concrete impos- 

 sible unless it is first treated. Generally, 

 prolonged washing with fresh water is 

 not sufficient, and frequent rinsing with 

 sea water is impossible when the distance 

 from the sea is great. The best technique 

 consists in covering the cement with a neu- 

 tral product : special paint, paraffin, plas- 

 tic substances, etc. In our installations, 

 paraffin has given the best results; it was 

 heated and spread with a brush and after- 

 wards impregnated by the flame of a blow- 

 torch. 



The lining of cement with plates of glass 

 is a very good method for large installa- 

 tions. Finally, the application of plastic 

 materials practically solves the problem. 

 At this time, there exist ready-made tanks 

 of 100 to 500 liters, of plastic that is abso- 

 lutely nonreactive to sea water. This tech- 

 nique is excellent. 



WATER CIRCULATION SYSTEM 



Once the tank has been installed, it is 

 essential that the breeding tanks, which 

 are generally located at a considerably 

 higher level, be supplied. At tlie present 

 time, there are rotary pumps made of plas- 

 tic which are immersed in the water and 

 operated by a shaft which plunges into 

 the water and is driven by an electric 

 motor. This device makes it possible to 

 supply a reserve tank which, by gravity, 

 supplies the breeding tanks. It is prefer- 

 able to any other device when a number of 

 persons use the installation. When the 

 reserve tank is full, it is sufficient to open 

 the desired faucet when water in one or 

 the other tank is needed. There is one 

 difficulty, however : the pump cannot op- 

 erate continously. It is therefore neces- 

 sary to provide an electric regulator, and 

 that introduces a possible cause of failure. 



Exhausting devices 



The system we have used is much less 

 cumbersome; it is based on water lifts 



which operate by means of air pressure, 

 viz., exhausting devices. These are ver- 

 tical pipes immersed into the water of the 

 reserve tank; their upper ends emerge in 

 the form of swannecks in the highest tank 

 of the installation. Air supplied by a 

 small electric pump is conveyed into the 

 tube at a sufficiently low level within the 

 reserve tank to move the water in the tube 

 and make it ascend into the upper tank. 



Our experience has led us to the formu- 

 lation of certain estimates which save 

 much guesswork. The interior diameter 

 of the main body of the exhausting device 

 must measure close to 6 mm. The pipe 

 may be rigid or flexible ; it may be of glass, 

 of vinyl, or of polyethylene. It may be 

 straight or bent, but must not have the 

 properties of a siphon. 



The air duct must open into the tube in 

 which the water ascends, below the level of 

 the water of the reserve tank, and it must 

 be moved closer to that water level as the 

 height of the exhausting device above the 

 reserve tank is increased (contrary to what 

 one might be tempted to do) . As a matter 

 of fact, the thrust of the ascending air 

 pulls a mass of air above the level of the 

 reserve tank, and that mass of air tends to 

 become smaller as the air arrives at a 

 higher level. It is necessary for the op- 

 eration of the apparatus that the water 

 mass above the level of the reserve tank 

 exercise a vertical downward pressure 

 from above, which must be less than the 

 upward thrust that the lieight of the water 

 in the reserve tank applies to the lower 

 part, of the pipe. In figure 1 we see the 

 exhausting device, represented by the pipe 

 AOB. ^6^ is the upper part and OB is 

 the lower part of the device; O' 0" is the 

 water level in the reserve tank. It is quite 

 evident that the influx of air at C will 

 drive a water column CO in the direction 

 of A. The farther C is from 0, i.e., from 

 the water surface, the larger will be the 

 column CO. As soon as this water column 



