Aiiicricaii Fisheries Society 89 



54; IStli, 56; 10th, 54; iOtli. 50; 21st, ':>1\ 22(1, 56; 23(1. 54; 

 24th, 54; 2f)th, 54. 



44ie eyes showed first on the 19th. six days after fertihza- 

 tion. It was difficult to detect them at this .stage owing to 

 the dark color of the eggs. Under the microscope the eyes 

 wlieii they first appeared were light colored but two or three 

 days later darkened considerably and were plainly visible. 

 The shape of the fish was plainly discernible on the 22d. al- 

 though the outline had been seen several days earlier. The 

 fry emerged from the eggs slowdy and instead of rising and 

 floating out into tlie trough through the lip, sank to the bot- 

 tom of the jar, where the}' rested cjuietl}- for se\eral days. 

 They were very dark in color and both in htie and general 

 appearance much resembled tadpoles. 



The percentage of hatch was \er}- small, only 400 fish 

 being" produced from the 6,000 eggs. ]Mr. Berkhous is of the 

 opinion that this small percentage was not due to failure to 

 fertilize but to cleansing the eggs of the glutinous matter in 

 the muck, b^-om my obserxations T am in accord with this 

 opinion, for the reason that vei'}- few dead eggs appeared 

 within the i)eriod when the}- should, and also from the fact 

 that they fungus.sed, and fungussed at a low water tem- 

 perature. 



Detection of dead eggs was very difficult on account of the 

 semi-opa(|ue character of the egg itself. It was only by 

 using a microscope that they could readily be distinguished. 

 Under a microscojje they show^ed a white opaque spot. 



One of tlie ])eculiarities of the work during the ])eriod of 

 incubatii)n was the impossibilit}- of siphoning the dead eggs. 

 Instead of becoming light and rising to the top the\- remained 

 mixed with the others and this ma}- also haxe had something 

 to do with the death of other li\'e eggs. It would ha\'e ])een 

 possible to se])arate tlie good from the bad onl\- b\- em]')t\-ing 

 the jar upon a tra}- and ])icking them off one bv one. The 

 ex])eriment as conducted indicates that better success in 

 hatching could be obtained b\- using tra}-s instead of iars. 

 Jn fhe liist })lace the eggs are so heavy that to move ihcm 



