162 BULLETIN OF THE UNITED STATES FISH COMMISSION. 



girdle is of mesoblastic origin, and appears just beneatli what may be 

 called the pectoral fold at the base of the breast flu. Its form and posi- 

 tion varies in different genera. 



13. The proctodeum or vent of the young fish appears long before the 

 stomodeum or mouth ; the intestine develops from behind forward, and 

 it is probable that the intestine and medullar}^ canal are primitively 

 continuous through the intermediation of a neurenteric canal. 



14. The number of somites or muscular segments varies in different 

 species at the time the tail is about to be formed, and is greatest iu 

 species in which a great number of myocommata are developed in the 

 adult, least in those in which the adult has fewer pairs of myocommata. 



PRACTICAL CONCLUSIONS. 



In the preceding account of the development of the Spanish mackerel, 

 it has been incidentally mentioned that the eggs of this fish are buoy- 

 ant, apparently from the presence of a large oil sphere in the vitellus, 

 and not because of the diminished specific gravity of the whole ovnm, 

 as apiiears to be the case with the cod egg. We would now insist upon 

 this character as being of such great physiological import that we can- 

 not aiford to ignore it or to conduct our hatching work without taking 

 cognizance of it in the construction of apparatus. The perfectly regular 

 development of the ova was found to take place practically at the sur- 

 face of the water, while those which sank to the bottom were considered, 

 in the light of experience, as not liable to develop at all. Where the 

 eggs were kept for the whole period of incubation in still water in a 

 marbleized i:)au, all that sank could be regarded as irrecoverably lost, 

 "while those which remained floating at the surface as uniformly hatched 

 ont. Changing the water on a lot of ova tbree or four times in twenty- 

 four hours in a pan gave almost as good results as any other method 

 employed. The active movement of the ova in apparatus devised to 

 hatch other species with heavy ova was amongst the least success- 

 ful modes, and especially where metals, such as copper, brass, tin, or 

 nickle, were used in the construction of the hatching vessels or screens. 

 Inasmuch as the protojilasm of the living egg is extraordinarily sensi- 

 tive to the poisonous effects of all metallic salts, such a result is no 

 more than might have been expected. We have therefore been con- 

 strained to suggest the use of apparatus which, as far as practicable, 

 was constructed of wood, glass, or of some material indifferent to the 

 action of sea-water. Experiments made, at the suggestion of Professor 

 Baird, with asphalt varnish and rubber paint at Cherrystone taught us 

 that it was possible to coat metallic surfaces with an inert substance 

 which would prevent the corrosion of the metallic vessels used in hatch- 

 ing, and hitherto found to be so fatally injurious to developing fish ova 

 of every kind. 



The i^crcentage of losses in every case was large, and I doubt if 25 



