12 



of water, to carry the work a stage further next season, 

 and also to try the same experiments in similar tanks at 

 the Piel (Eoa Island) laboratory. 



Last April we fitted up, in the lower floor of the 

 Aquarimu house at Port Erin, three wooden hatching 

 tanks, each 5 feet by 3 feet by 1 foot, and so arranged, like 

 steps, that water could flow by bamboo spouts covered 

 with a fine silk net through the series of tanks. From 

 the lowest wooden tank the water fell into a concrete 

 floor tank, into which dipped an endless chain formed of 

 an india-rubber belt bearing numerous little buckets. This 

 chain of buckets revolved on a drum, octagonal in section, 

 which was kept in motion by india-rubber belting passing 

 from its axle to a pulley on a large water-wheel actuated 

 by the fresh water tap (see PI. I.).* 



Consequently, by turning the tap the whole apparatus 

 was set in motion, and the sea- water from the concrete 

 floor tank was raised by the little buckets and emptied 

 into a sloping wooden trough which guided it to the upper 

 hatching tank. Thus the same water was used over again, 

 a couple of gallons of fresh sea- water being added to the 

 system every day. 



During the period when the apparatus was working the 

 temperature and the specific gravity of the water in the 

 tanks kept fairly constant, the extremes of the range 

 being : — 



Temperature from 50^ to 53*^ F., and 



Specific gravity from 0-0265 to 0-0270. 



Each of the three tanks had a partition 1 foot from its 

 outflow end which stopped 2 inches from the top, and a 

 second partition 6 inches nearer the end which reached 



*Tlic tanks and tlie water motor apparatus were made most carefully and 

 ingeniously, from our plans, by Mr. R. Garner, superintendent of the wood- 

 working department at University College, Livei'pool. 



