Bulletin of the united states fish commission. 163 



per cent, of tlie whole luunber of eggs was ever hatched out even under 

 the most favorable conditions. The utility of some cheap and effective 

 glass ai)[)aratus is very apparent from experiments made by Colonel 

 McDonald, as his system admits of a wide range of application. Other 

 methods, especially those in which the intermittently active siphon 

 principle was applied, seem to afford some promise that a successful 

 iipi)aratus may yet be devised to work on that plan. Some trials of 

 such apparatus made by Colonel McDonald were promising, but I leave 

 the results attained for him to report upon. An equally simple hatch- 

 ing box was extemporized by Mr. Sauerhoff with a Ferguson cylinder 

 set into a tub, the eggs being placed in the cylinder and the constant 

 water supply allowed to escape through its sides and bottom into the 

 tub outside and run off by a proper outlet. With this apparatus a fair 

 degree of success was obtained. 



It may be stated as a general principle that buoyant ova must have 

 gentle treatment. That if they are much agitated in the water they 

 tend to be injured and are carried to the bottom, where they die. It 

 appeared that when the normal buoyant tendencies of the ova were in- 

 terfered with by any of our methods large losses were the result, and 

 that the nearer our methods approached the natural environment of 

 naturally spawned ova in the open sea the more successful we were. 

 To lorcibly immerse the egg of the mackerel, and keep it immersed, 

 would simply be to thwart what is most palpably a normal condition of 

 its life at the surface of the water. 



The fertility of the mackerel, like that of most fishes with floating 

 ova, is very great, and it is to be expected that the mortality t)f the 

 ova will be in proportion to the fertility of the species. This seems to 

 be borne out by the converse state of affairs in the stickleback and top- 

 minnow, where a small number of embryos are matured, but are de- 

 veloped, with little or no losses, in a nest, and are nursed by the male 

 or viviparously in the body of the female. Viewed in this light, we 

 should expect large losses in hatching the cod, Spanish mackerel, moon- 

 fish, and bonito or crab-eater. 



The young fish began to feed on the fourth day after development had 

 begun, and on the third day after hatching. The true nature of the 

 food was not determined, as it was seen in the intestines of only a few 

 specimens. Inasmuch as the young fish by the end of the first week 

 of their life already have teeth, it is easy to believe that their food con- 

 sists almost entirely of small articulate animals, which, with their quick 

 darting movements, they might readily single out and capture while 

 swimming about in water in which such prey abounded. 



