352 H. M. Vernon 



thought possible that a combinatiou of small quaiitities of green and 

 red weeds together might bave a favourable effect upon the growth 

 of the larvai. In Expts. 68, 133 and 42 are given the results obtained. 

 Here it will be seen that in oue case the larvie bave increased 3.6^ 

 in size, and in the other two sligbtly dimished. The percentage 

 number of ova reaching the larvai stage is, however, greater than the 

 normal, and hence on the whole the eftect appears favourable, in 

 spite of the ammonia present having as a rule increased. In the 

 last experimeut in the table, made with the branching green weed 

 Dictyota^ the larv;e were 3.1^ snialler than the normal, probably 

 owing to the fact that the alga had got into an unhealthy condition 

 towards the end of the experiment. 



The observatious on red weeds were not continued, because of 

 the difficulty of getting them to live in good condition. In some cases, 

 as in experiments 42, 90, 92 and 133, pieces of weed attached to 

 stones were obtained, bat this seemed to make no difference. In the 

 case of Ulva^ the alga appeared to remain equally healthy, at any 

 rate for eight days, whether it was attached to a stone or not. Thus 

 in Expts. 39, 41 and 42 pieces of the alga growing from stones, and 

 in the other experiments torn off pieces, were used. 



A compari sou of the effects of green weeds such as Ulva^ and 

 of red weeds such as GeUdium, upon the purification of the water 

 and the growth of larva;, obviously speaks much in favour of the 

 Ulva. Thus this alga rapidly removes the free ammonia present in 

 the water, and probably under certain conditions may diminish the 

 organic ammonia. Also it seems to keep in a healthy condition 

 much more readily. Red weeds on the other band, though they seem 

 in some instances to exert a favourable influence on the growth of 

 the larvic, increase the ammonia present in the water, and are 

 moreover exceedingly diftìcult to keep in a healthy condition. 



The Effects of Diatomace« and minute Alg?e. 



a. The Effect of Filtration through Sand. 



Filtration through sand is the method most frequently adopted 

 for the purification of fresh water for drinking purposes. Until com- 

 paratively recently, this piirifying action was looked upon as a 

 raechanical one, but K. Kocii has shown' that it must be attributed 



1 Zeit. Hyg. 14. Bd. pag. 395. 



