[9] WOKK AT COLD SPRING HAEBOR. 137 



we sball need to run it ouly about six hours per day when we are using 

 the water. 



The large pond was found to require banking outside the wall as well 

 as inside, for the seas went through it and eut out the inside embank- 

 ment. In this place we were fortunate in not striking springs of fresh 

 water, which are common all along the beach, but we found great 

 trouble from this cause, and also from quicksands, in laying the drain- 

 tile. The winter of 1883 closed in early and we were compelled to sus- 

 pend out-door labor and to defer until spring the completion of the great 

 tidal reservoir, but we were enabled to hold the water as high as half 

 tide and to begin work. The hot-air engine worked very well, and we 

 hatched the eggs of some fish which laid them in clusters on the sea-weed 

 and which the fishermen all erroneously declared to be those of the little 

 tom-cod [Microgadiis tomcodus)^ locally known as "frost-fish "in the fall 

 of the year, and as tom-cod in the spring. I sent some of these eggs to 

 Prof. J. A. E^'der, at the central hatching station of the United States 

 Fish Commission, and he hatched them in artificial sea-water. The 

 spawning season of this unknown fish is in November and December, 

 and they had finished spawning before our engine was in position, but 

 we gathered the eggs from the seaweed, to which they are attached in 

 bunches the size of a hen's ^gg, and are easily obtained by the oyster- 

 men when raking for oysters. What the eggs were I will not attempt 

 to guess, but the following year, 1884, we took the eggs of the tom-cod 

 from the fish and found them free and heavy, and the appearance of the 

 embryo differed from the unknown eggs. 



In the winter of lSS3-'84 we obtained several million codfish eggs 

 from the cars at Fulton Market, but none of them were good. They 

 showed the shrunk vitellus which gives both them and shad eggs a 

 "speckled" appearance, which indicates that there is no possibility of 

 impregnating such eggs. In every case the parent fish had been 

 brought in the well of a fishing smack, and after being dipped out had 

 been thrown into the floating car alongside, falling from 4 to feet, 

 usually on the abdomen. This, in my oi)inion, is more than the delicate 

 cod egg can stand. The membrane, or shell, covering the egg of the 

 codfish, is so delicate that a light touch of the finger, when the egg is 

 on any hard substance, will burst it like a soap-bubble, while a trout's 

 Qgg will bear the hardest squeeze that can be given between the finger 

 and thumb. In December, 1884, we obtained one lot of eggs from the 

 same place, which floated and appeared good when taken, but w^ere 

 dead nnd at the bottom of the jars on arriving at the hatchery. Later 

 the fish came in dead and we have never had good cod eggs in the 

 hatchery. In ^^Tovember, 1884, 1 visited Wood's Holl and saw the a])pa- 

 ratus devised by Capt. II. C. Chester, and on my return made a similar 

 one of zinc instead of wood. I used a big wooden tank 12 feet long by 

 (3 feet wide and 3 feet deep. The hatcher was merely an elliptical 

 piece of zinc, 5 feet 8 inches long by 2 feet wide, without bottom, and 



