1897J IMICROSrOPICAL JOURNAL ll'J 



water, eiv., arc all distiirliing elements aH'ectiiig their 

 growth. 



Ill order to deterniine the effect of light upon their 

 growth it was found necessary to make experiments in 

 the open reservoirs under conditions practically the same 

 as tliose found in nature. 



The method employed was an extremely simple one. 

 It consisted of suspending- bottles tilled with water from 

 the same source at different depths in Ihe pond, the bot- 

 tles being- tied to a ro})e which hung from an anchored 

 buoy. After a certain time the bottles were drawn to 

 the surface and the water examined, records being kept 

 of the number of diatoms in each sain})le before and after 

 exposure. The bottles varied in capacity from 150 to 

 1,000 cc. In the first Hve exjieriments they were tightly 

 stoppered, but in the later ones silk bolting cloth was 

 tied over the mouths of the bottles, and inverted glass 

 tumblers were })laced above. The latter arrangement 

 gave much heavier growths on account of providing bet- 

 ter op[tortunity for the cirt^ulation of air and for the re- 

 newal of food supply. 



Without descril)ing the experiments of [Forel Forel, 

 F. A. "Le Leman, niouographie limiiologique," Laus- 

 anne, 18951 and others upon the intensity of light at 

 various depths, it may be said that the decrease in the 

 intensity below the surface is due to two causes — ab- 

 sorption by the water, and the |)r'^sence of fine particles 

 which act as a screen. The reduction of light in passing 

 through water is supposed to follow the law that as the 

 depth increases arithmetically the intensity of the light 

 decreases geometrically. For examjile, if the intensity 

 of the light falling u[)on the surface of a pond is repre- 

 sented by 1, and if \ of the light is absorbed by the first 

 foot of water, then the intensity of light at the depth of 

 one foot will be i}': the second foot of water will ab- 

 sorb i (»f ii, and the intensity at a depth of two 



