114 



SEASON 1874-75. 



to match, one to use at Middlethird and the other for Craigend 

 hatching-room. The bottom of each was the same size as that of 

 a slate tray, but the ledge was placed on the sides instead of the 

 ends. They were made 9 inches deep inside, and divided by two 



divisions into two compartments 

 of 7^ inches and one of 3 inches 

 (Fig. 58). Two frames of wood, 

 well charred, were covered with 

 wire-cloth and rested on the 

 ledges 1 inch above the bottom ; 

 the two compartments were then filled with washed and boiled 

 gravel to within 3 inches of the top. Three holes are bored, one 

 in each compartment, to facilitate cleaning, and a hole near the top 

 in the last division admits the filtered water to the feeding- 

 spout (Fig. 59). The division between the two filtering compart- 

 ments is stopped 1 inch from the 

 bottom. The action is as fol- 

 lows : The water is received 

 into the first compartment, either 

 from the cistern, as at Craigend, 

 or from the house-filter, as at 

 Middlethird ; in both cases the 

 supply is regulated above the slate filter, which must be regarded 

 as an extra purifier merely. The water falls through the gravel, 

 and passes freely into the centre compartment, where it rises 

 through the wire-cloth and gravel, and falls into the small com- 

 partment, from which it passes to the trays (Fig. 60). The advan- 

 tage of this filter is the very large filtering area it 

 possesses compared to the space occupied and the 

 quantity of gravel required. By removing the 

 corks the filter can be cleaned in a minute if neces- 

 sary. The only objection I found was its liability 

 to freeze at Middlethird, where it was subse- 



FIG. 60 scale & 



quently discarded, and an extra slate tray sub- 

 stituted ; but at Middlethird I had placed a 



FIG. 59 scale ^. 



