[3] APPARATUS FOR HATCHING SPANISH MACKEREL. 1097 



aquaria. The overflow was established by meaus of a syplion consist- 

 iug of a glass tube that was inside the aquaria aud a rubber tube that 

 fell over the side and reached below the aquaria ; to the lower end of the 

 glass exit tube we affixed a strainer consisting of a wooden cone, and 

 over the 3-inch opening fastened a plate ol" mica i)erforated by niinnte 

 holes. We watched its operations with interest, and at first found it 

 was all our crude ideas anticipated, and although we used a strong cur- 

 rent of water, yet the floating eggs were not in the least disturbed. But 

 after remaining in the water for some time the mica became very brittle 

 and cracked. This would not do, so we substituted for the mica perfo- 

 rated tin coated with asphaltum, but the asphaltum did not coat the per- 

 forations, and these soon corroded, and the oxide of iron excluded the 

 openings. 



Later, at Mr. Smith's suggestion, we tried silk gauze, such as is used 

 in making nets for surface towing, and by doubling this over the ends 

 of the wooden cones preven.ted the escape of the eggs, but our trouble 

 lay in the dead eggs rising and becoming entangled in the meshes of the 

 gauze, thus i)i"eventiug a free current, but this difficulty was easily over- 

 come by occasionally freeing the surface with a feather. 



The eggs hatched well, fully 60 per cent, of young fish appeared, and 

 in 17^ hours after impregnation, with the temperature of the water at 

 80° F., a microscopic examination led me to believe had all the eggs 

 placed in the aquaria been properly impregnated the percentage hatched 

 would have been greater, but in taking the eggs it is impossible not to 

 get some unripe ones. After hatching, the fish gradually died, and for 

 a long time the cause was unexplained ; finally we decided it must lay 

 in the iron tanks used as receptacles for water. These tanks have been 

 used to hold fresh and salt water for three years, and, though repeatedly 

 scraped and painted, soon corrode; besides what now must have been 

 suspended in the water we found large quantities, comparativel.> speak- 

 ing, deposited in the bottom of a vessel where some of the water stood. 



We had no good opportunity to test our conclusions, for later, when 

 Captain Wood substituted a cypress tub as a receptacle, the cold nights 

 destroyed both eggs and fish. Captain Wood also changed, in these 

 later experiments, the form of apparatus by immersing into a vessel con- 

 taining water a glass cylinder over whose lower end gauze was stretched ; 

 into this cylinder the eggs were placed. The entrance water either fell 

 into the vessel outside the cylinder and fresh water was obtained by 

 dift'asion through the gauze or the water fell into the cylinder direct 

 and passed out through the gauze; but in either case the eggs which 

 settled caused much inconvenience by covering the gauze and prevent- 

 ing the free circulation of water. The cold, however, cut short our work. 



This is a brief synopsis of my own observations, and although in the 

 main our work was not a success, yet there appears to me evidence that 

 with slight modifications our original apparatus will hatch the eggs of 



