ARTIFICIAL PROPAGATION 183 



between the standards are filled in with sawdust to help to preserve an 

 even temperature. 



The floor space, sloping towards the front door, is occupied by 

 eleven hatching boxes, which, as the plan shows, are arranged to allow 

 of convenient working. Against the left wall are two boxes end to 

 end. Two groups of four boxes each, arranged end to end in pairs, 

 occupy the central floor space, while against the right wall, away from 

 the door, is one box. The space left on this side near the door is useful 

 for tackle. 



The hatching boxes stand on frames at a convenient height from 

 the floor, and, as at Luss the water supply enters from the rear, the tier 

 of six boxes furthest from the door is on a higher level by the depth of 

 a box than the tier of five nearest it, each upper box draining into its 

 corresponding box on the lower level. 



To trace the water system more precisely, the supply is delivered 

 from a continuation of the main -^ in. lead pipe within the hatchery to 

 each of the six upper boxes. Separate connections with this pipe are 

 made for each box, the taps being finished with f in. nose-cocks. After 

 serving the upper boxes the water flows over lipped slots to the lower 

 tier, whence each box drains to a common lead waste-pipe carried 

 beneath the flooring through the doorway to a fireclay drain-pipe 

 leading to the river a few paces away. 



The supply of the lower tier of central boxes is, at Luss, by a 

 separate connection with the main inside the hatchery, supplemented 

 by spray water for the reason I shall indicate when describing the 

 hatching system. I need only add at the moment that the whole water- 

 supply system within doors permits of each box being worked as an 

 independent unit. 



There are two systems of hatching the ova in vogue at Luss, and I 

 shall discuss the respective merits of both, namely, the glass grille 

 system and the perforated zinc basket system, which has, I believe. 



