12 Transactions of the Society. 



the cut ends, streams of bubbles will emerge when the apparatus 

 is working. A bit of dried hawthorn as thick as one's little finger, 

 which has been dead and exposed to the weather for a year or 

 more, gives an extremely fine division of the air. But most pur- 

 poses will be served quite well by woods with coarser vessels. 

 Insufficient previous weathering is apt to give trouble through 

 swelling taking place after immersion. An air valve may with 

 advantage be set in the main stem of the rubl ler tubing in order 

 to obviate all tendency to reflux, or better still, a valved extensible 

 air reservoir with suitable recoil may be interposed, thus eliminating 

 altogether intermittence of aeration. It goes without saying, also, 

 that special arrangements may be added, such as those which 

 Browne * has so successfully devised for the growth of hydroids. 



To sum up, the influx of water through A is constant, and the 

 sequence of events is as follows : — Base of pressure in C ; rise of 

 water in E and F ; forcing of air through the aerating nozzles ; 

 overflow of surplus water through G ; filling up of C ; commence- 

 ment of siphon action ; flowing back of water from F into C ; 

 entrance of air into C ; emptying of water into C ; cessation of 

 siphon action ; recommencement of rise of pressure in C. 



The apparatus can be fitted up wherever there is a constant 

 water supply fresh or salt, under even a slight degree of pressure. 

 It works quite automatically, and after being properly adjusted 

 requires no attention except in arranging the nozzles from time to 

 time to suit new aquaria, or the varying requirements of aquaria 

 already established. No active damage can be done to the aquaria 

 even though the siphon action from any cause (e.g. leakage or 

 slackening of inflow) should temporarily cease. 



The air used is freed from most of its soluble and suspended 

 impurities through entering by the long wet tube F, bubbling up 

 through the water in B, and remaining for a time within C, into 

 which there is a constant inflow of water. By way of further 

 precaution, the air entering at M may be filtered through cotton 

 wool. Thus the atmosphere even of a city laboratory, may be 

 rendered practically harmless. 



The intervals of pause and of active aeration can be varied 

 within wide limits, as also can the periodic time of both. Or the 

 whole apparatus can be stopped and re-started after a time (e.g. 

 to imitate tidal conditions) by turning the water-tap. 



A single aerator on the scale indicated above will provide con- 

 tinuous and efficient aeration for as many as thirty small aquaria. 

 Increase in the sizes of the parts gives an increased supply of air, 

 while by heightening the various tubes the internal pressure can 

 be brought up almost to the limit of pressure of the water supply. 



* Journ. Marine Biol. Assoc. United Kingdom, rt.s. viii. No. 1, pp. 37-43. 



