The Relations between Marine Animai and Vegetable Life. 361 



effected, and the ,physiological' purity. lu Expts. 100 and 124, in 

 whicb the free ammonia in the water was largely increased in aniount, 

 the larva3 were nevertheless appreciably increased in size, whilst in 

 Expts. 99 and 123, in which it was very largely redueed, the eifect 

 was practically nil. We can only conelude, froni this want of nni- 

 formity in result, that there is another factor present, which thus far 

 has not been taken into account. This factor, as will presently be 

 made evident, is in ali probability of a bacterial nature. 



b. The Effect of keeping the Water in Diffuse Daylight. 



As has been shown above, a vegetable growth is gradually 

 deposited on sand through which a stream of water is allowed to 

 fìlter. It follows therefore that the water itself contains small 

 quantities of algae and diatoms in Suspension. In order to determine 

 something further as to the purifying capacities of these minute 

 quantities of vegetable matter upon the water, and as to the rapiditj: 

 with which they can act, a large covered jar of the ordiuary Aquarium 

 water was allowed to stand in a room where it was exposed to no 

 direct sunlight. The chemical analyses made from time to time gave 

 the following results. 



Free NH3 Organic NH3 



Originai water .223 mgm. .134 mgm. 



After 10 days .134 .171 



- 25 - .017 .269 



- 33 - .010 .227 



These fìgures show that the free ammonia present in the water 

 is very rapidly removed by some agency or other, only about \% 

 of it being left after 33 days exposure to diffuse light. The organic 

 ammonia on the otber band is considerably increased, it being double 

 its originai amount after 25 days exposure. That these striking 

 variations in the composition of the water are due chiefly, if not 

 entirely, to a vegetable agency, is rendered probable by the fact that 

 after the water had been standing a few days, small patches of 

 vegetable growth, some of them bright green, and others purple 

 brown in colour, began to forni on the sides of the vessel, and these 

 patches steadily increased in size as the experiment continued. On 

 microscopical examination, this growth did not appear to be of a de- 

 finite character, but in the deposit at the bottom of the jar there were 

 a few remains of diatoms. As the jar containiug the water had been 



