Mr George Brook on the Aeration of Marine Aqiiaria. 385 



pump would not do its work properly. Another item which 

 caused me no slight alarm was the great price which the 

 pump cost to work. 



In order to pump 850 gallons of sea water, it was necessary 

 to use 1250 gallons of fresh water as the motive power, so 

 that under the most favourable conditions 33% of power was 

 lost, and in fact at many times a great deal more. Altogether 

 I became dissatisfied with the arrangement, and appealed to 

 my friend Mr J. T. Nelson of Leeds to make me a plan of a 

 pump that would prove more efficient. The result was, that 

 I had made a double-action lift and drive pump, placed 

 horizontally so as not to lose its prime. The general con- 

 struction of this pump will easily be understood on reference 

 to Fig. 2. A is an accurately bored glass cylinder 12 in. 

 long and 4 in. in diameter, obtained from Messrs Chedgey. 

 The piston rod is of steel, coated with ebonite, and afterwards 

 turned and polished so as to work smoothly. The piston 

 itself is made of lignum ritm, and has a round indiarubber 

 ring in its centre so as to form perfect contact. The ends B 

 and C are also made of lignum vitce, and the valves in them 

 are round indiarubber balls weighted inside with lead. The 

 dotted line above the valve box at B and C indicates a groove 

 cut out in the middle, so that, in case the valve should stick, 

 the water can still pass through. The pipes E and E' are in 

 this case cast-iron lined with glass, and have an internal 

 diameter of 2 in. The whole is bolted to a stout plank, D, 

 and connected with the shaft of the water motor in the 

 ordinary manner. It will be easily seen from the figure that 

 a stroke of the piston in either direction both fills the 

 cylinder from below, and drives the water already in the 

 cylinder into the fall-tank. This pump has now been in 

 work over twelve months, and has given every satisfaction. It 

 requires very little attention, and is not liable to get out of 

 order. With a rotary pump a certain force has to be main- 

 tained in order to lift the water the required height. Thus, 

 if the motor should be working below its normal speed, no 

 water is raised by the pump, and the power is wasted. The 

 case is quite different with the pump here described ; all the 

 . motions being positive, a minimum of power is lost, and 



