MIDDLETHIKD HATCHING-HOUSE. 



133 



and a splash-bar separating the two. The water fell on the zinc, 

 and became a shower of spray. 



This inverts the old myth the recipient fry became golden, 

 the shower remained soft water. All the rearing-boxes were 

 fitted with safety-screens, and the water was heightened by means 

 of a strip of coarse flannel laid across the outlet. Below is a 

 section of the rearing-box (Fig. 87). 



FIG. 87-scaIe 



In arranging the internal fitting I was guided by a desire to 

 be able to show everything off to the best advantage. The whole 

 thing was yet a toy, and the problem was how to walk round it. 

 It was to this end the E-spout was adopted ; without it I could 

 not have fed the rearing-boxes, except by a spout across one of the 

 passages. The arrangement will be best understood from the plan. 

 The water from the outside filter was conveyed in a covered spout 

 to the outside of the hatching-house. This spout was carried on 

 a dike to the level of the top of the inside filter. Just outside 

 the water was divided, what was required for the filter passing a 

 sluice which set back against a slight head 1^ inch, if I remember 

 rightly, and all the rest of the water fell into the upper end of the 

 E-shaped spout, through which it passed into the distributing- 

 spout, in the end of which was an overflow, so as to keep the head 

 constant on the overflow sluices feeding the rearing-boxes. The 

 overflow at the end of the distributing-spout was made half an inch 

 higher than the others. This was in practice found sufficient. The 

 water passing the end overflow is an apt illustration of the theory 

 of waste. So long as a drop passes, so long will every other spout. 

 sluice, and tap run exactly as intended, so long, and no longer. 



This overflow waste maintaining the pressure on the tap that 



