91 6 BULLETIN OK THE BUREAU OF FISHERIES. 



of all the gravel and all the sand, the layers of which lay flat on an area of 2.4 

 meters by 0.6 meter. 



The filter, in short, was not sufficiently cleaned, the water was turned on 

 again, and not sufficient note was made of the fact that it was the 8th of May 

 and the river water at a temperature of 17°, the first result of which would be 

 a sudden mortality among the trout. This happened, too. From May 8 to 

 May 1 7 the water was maintained at this temperature, and on the morning of 

 the last day I noticed some 30 trout dead in the pond, while the others were 

 being carried to the grating at the outlet. Most of the fish turned over and over 

 and made pirouettes, then jumped into the mass of water as if to cross it in one 

 movement, but they fell exhausted and dropped to the bottom. About 11 

 o'clock in the morning 300 trout were dead. 



The fish had "the staggers." In two days they grew somewhat dark, then 

 they began to weaken, swam with difficulty, and could no more maintain a 

 horizontal position; their 'behavior was abnormal, and they finally died. All 

 of the trout did not show the weakening, but fell to the bottom of the pond 

 and succumbed after opening their gills convulsively two or three times. 



I gathered the dead trout with a net and counted 1,200 of them. That 

 evening there were more on the bottom of the pond, and by the evening of the 

 next day the mortality was almost complete. The remainder living was an 

 insignificant number. I believe that 160 were placed in spring water, and out 

 of these 160 only about 100 were saved. 



This mortality had been caused by the imperfect action of the gravel filter 

 and by its dirtiness, by the fermentation of the organic matter which the filter 

 had retained when from April 24 the temperature of the water, 12°, rose to 

 17° about May 10. The cleaning of the upper layers of fine sand of the filter 

 gave free passage to the Saprolegnia, the mycelium of which abounded in the 

 lower parts or layers of gravel, and all this infection invaded the basin, with the 

 effect of all as just described. 



In order to remedy this at least partially, if there is no spring water at 

 one's disposition, it is better not to use any filter than to use such as this. The 

 great quantity of water necessary in a fish-culture establishment makes it 

 difficult to obtain a perfect filter. It might be well to use water from a river, 

 easily aerated if the pond is lower than the river. 



The best means of doing without a filter would be to keep a small fry 

 trough very clean, so as to be susceptible only to disease or an infection coming 

 from without, which is much at best. In order to attain this practical result, 

 a siphon-outlet system, of a kind such as I have devised (op. cit.), should be 

 installed. 



