286 PURIFICATION OF PUBLIC WATER SUPPLIES. 



ing a matter of large cost. Some of the European beds 

 are covered. So far as known, the few built in this 

 country are open, except one or two of very small area. 



The difficulties mentioned arise from three principal 

 causes, namely, algse growths on the sand, "conditions of 

 weather preventing cleaning, and changes in the charac- 

 ter of the water supplied to the filter. 



The conditions vary in different localities. Their 

 effect is more marked in our own case for the reason that 

 when our beds are in their best possible order they are 

 too small in area for the work required of them, as shown 

 by the foregoing comparisons. 



With us, in warm weather, commencing sometimes 

 with the latter part of May, the development of green algse 

 begins in small patches on the surface of the sand. 

 Under favorable conditions it spreads rapidly and soon 

 covers the entire area and stops filtration as effectually 

 as a rubber blanket. It cannot be removed so long as 

 any water remains above it. Although forming an 

 impervious covering when undisturbed, so soon as it is 

 touched with a shovel or other implement for its removal, 

 if the water is still over it, it flies in all directions in 

 minute pieces, only to settle in a short time and com- 

 mence growth as vigorously as ever. 



Therefore the water must be entirely drained from the 

 surface of the sand before its removal can be effected. 



In our case, as is usual, there is no provision for com- 

 plete draining except down through the sand. When- 

 ever, therefore, the development has gone so far as to en- 

 tirely, or nearly so, stop the passage of water, a long 

 time will be required to drain the sand sufficiently to 

 permit it to be jjroperly cleaned. If at such times the 

 water in the reservoir should be low there is great danger 

 that the supply of filtered water shall be exhausted be- 

 fore the beds can be put in operation. Of course it is 

 the constant effort to maintain a suflScient quantity of 



