358 H. M. Vernon 



removal of suspended organic matter. After a week's filtration 

 however, the sand had becorae impregnated with enougb vegetable 

 growth to remove a quarter of the free and an eigbth of the organic 

 ammonia, whilst after a fortnight three quarters of the free and a 

 third of the organic ammonia were absorbed. As in the previous 

 case, the cessation of the current of water for twenty-four hours on 

 the 22"'^' day of the experiment caused a great diminution in the 

 absorption power of the sand for organic ammonia, though the free 

 ammonia was removed more efficiently than betöre. The rate of 

 filtration of the water was also diminished more than threefold. On 

 the 29*'' day of the experiment, after collecting and analysing a 

 specimen of the water, the top half centimetre of the sand was 

 scraped away, and the outflowing water again analysed. It now 

 fìltered tbrougb tbree times more rapidly than before, but was just 

 as efficiently purified of its free ammonia. Another layer of sand, 

 two centimetres in depth, was then removed, and the water again 

 analysed. Its purifying action on the free ammonia was now 

 considerably less than before, but this may bave partly been due to 

 the greater rapidity of the filtration. As in the previous series of 

 observations, this continued purifying capacity of the sand could 

 scarcely bave been entirely due to the layer of diatoms and algte 

 with which the sand exposed to the light at the sides of the bell-jar 

 was impregnated. Bacteria or other organisms deposited in the sand 

 itself must bave effected some of it. As before, the effect of diminishiug 

 the rate of filtration was also determined. In that, owiug to the 

 presence of much smaller quantities of vegetable growth, the purifying 

 capacity of this filter was much less, so one would expect, on a 

 slowing of the current, that the increase in the amount of ammonia 

 added to the water by the decomposition taking place in the vegetable 

 growth would also be smaller. Such is actually the case. Thus on 

 the eighth day of the experiment, with a very slow rate of filtration, 

 the free ammonia was increased by only 33^; and on the 15*'' day, 

 by only 76^, instead of the 200^ and upwards observed in the 

 former case. 



In addition to testing the purification of the water by chemical 

 means, it was tested physiologically by observing its efifects on the 

 growth of larvse. The following were the results obtained. 



