284 BULLETIN OF THE UNITED STATES FISH COMMISSION. 



"witli filaments, but lie tliouglit the fibers were inside the egg-membrane 

 instead ol' outside of it, as may be learned by reference to his paper on 

 the subject in AlUller^s Archiv for 1855. Professor Kolliker, in the Ver- 

 handl. d. 2)hysik. u. med. Gesellschaft in Wiirzhurgj eighth volume, for 

 1858, rectifies Ilaeckel's observations, and shows that the fibers or fila- 

 ments are external, but it does not appear that he ever understood their 

 real function, viz, to provide a means of fixation and support while the 

 ova were undergoing incubation. Haeckel unfortunately observed only 

 uurii^e ova, contrary to what he supposed, as is clearlj- shown by his 

 figures, but he found the fibers present in the eggs of Belone, Scomber- 

 esox, Hemirhamphus, and Exocoetus. The writer has observed them in 

 the unripe eggs of Hemirhamphus iinifasciatus, but has not had any op- 

 portunity to observe them in the egga of the tlying-fish, Exocoetus. In 

 all of these the fibers are distributed and attached at intervals over the 

 whole surface of the egg, but in the Atherinidce, as shown by my ob- 

 servations on the eggs of Chirostoma notata in Mobjack Bay, Virginia, 

 in 1880, there are only four filaments, which are attached to the 

 vitelline membrane at one i>ole of the egg and quite close together. 

 These are at first coiled around the vitelline membrane in one i)Iane 

 quite closely, as in Belone, but they unwind when the eggs are dis- 

 charged into the water, when the threads of adjacent eggs become en- 

 tangled so as to form clusters of considerable size. In this genus the 

 filaments are nearly half an inch long, without a swollen base, attached 

 to an. egg one-sixteenth of an inch in diameter. We saw that all of the 

 genera of Scomheresocidoe were found to have their ova provided with 

 filaments. The genus Arrhamphus is the only one the eggs of which 

 have not been observed. It is very probalile that all of the genera of 

 Atherinidw have ova with filamentous processes ; at any rate it is desir- 

 able that they should be looked for in Atherina, Athennichtliys, Teira- 

 gonurus, and Labidesthcs. This sui)position also raises the question 

 whether the MugUidce do not have eggs of the same kind. A confirma- 

 tion of this hypothesis would be desirable in that the large number of 

 species in the family, their wide distribution and considerable size, con- 

 spire to render them of value as food-fishes over a large area of the 

 earth's surface. It is quite as important for us to know what natural 

 means exist to favor the survival of the germs of species as it is for us 

 to know what artificial means to provide for their increase and i>rotec- 

 tion. In fact the latter kind of knowledge ought to be based upon and 

 sui)plement the former, since it is by a combination of natural ami arti- 

 ficial protective agencies that much more can be done to increase the 

 number of food-fishes than by the latter alone. It may indeed happen 

 that we Avill yet learn that certain species need no protection save that 

 which would prevent their capture during the breeding season. 



DEVELOPMENT. 



The development of the germinal disk of the silver gar is essentially 

 like that of the Spanish mackerel and the cod. The egg is quite large, 



