34 THE NATURAL HISTORY AND HABITS 



double the thickness it was when it appeared first ; 

 the shell of the egg appears more thin and crystalline, 

 so that we can see all parts of it perfectly, and can 

 examine it more minutely. When seventy-nine days 

 among the gravel, the above appearances have in- 

 creased a good deal. The red of the yolk is less, and 

 the eyes of the fish larger, the shell of the egg is more 

 thin, and the young fish larger and filling more of the 

 inward space, and easier seen. At ninety-three days, 

 the shell looks a good deal distended, the head of the 

 fish larger and better formed, the eyes seen more per- 

 fectly, and the red yolk leaving the back part, and 

 drawing more closely towards the belly of the fish, 

 where it forms into a conical looking bag, one end of 

 which is attached to the belly of the fish below the 

 gills. This appears yet more clearly at one hundred 

 days deposited, far now we find the young fish alive 

 within the shell ; we see it turning over from one side 

 to the other, and the shell getting very thin and dis- 

 tended ; the above red appearance in the yolk is now 

 fully formed into the conical bag, and attached to the 

 neck and belly of the young fish, and when we see 

 the fish turning over within the shell, we observe this 

 bag turning over also. At this time, by taking a 

 grain of seed along with a little water into the hollow 

 of the hand, the heat of the hand raises the tempera- 

 ture of the water, which causes an immediate move- 

 ment of the fish, whereby the whole formation in all 

 its different parts are clearly seen, and sufficient time 

 got to observe all the various alterations and progress, 

 which at this time are very interesting. We now see 

 the new generation of fishes all alive, although yet 

 confined in their original cells. When we consider 

 that these yet unshapely things, not more than half an 

 inch long, will in eighteen months hence afford sport, 

 and fill the basket of the angler, as well as the box of 



