286 BULLETIN OF THE UNITED STATES FISH COMMISSION. 



inasmuch as tbe vesicles in the germinal pellicle disappear when the lat- 

 ter has been aggregated into the germinal disk, may it not be that they 

 represent the fragments of the disintegrated nnclens? This view, how- 

 ever, as already stated in my paper on the Spanish maclierel, is nega- 

 tived by the results obtained in staining the germinal i)ellicle of the cod 

 eggj where these vesicles remain untinged. No oil spherules are visi- 

 V)le in the vitellus, the latter being optically homogeneous. The whole 

 egg is heavier than the sea- water, and quickly sinks to the bottom ; its. 

 specific gravity must therefore be much greater than that of the shad or 

 salmon. 



The germinal disk is developed in the usual way by the aggregation 

 of the germinal i^rotoiilasm of the pellicle, which covers the vitellus at 

 one pole of the latter. It does not appear that impregnation certainly 

 takes place before the formation of the germinal disk. Observations on 

 this point are, however, still too scanty and untrustworthy to be of 

 much value, and until special attention is directed towards tbis point 

 it will be most commerulable to maintain a skeptical silence in regard 

 to the views held on this subject. Special apparatus is needed to con- 

 duct researches on the phenomena of impregnation of fish ova, supple- 

 mented by reagents which will act quickly, so as to fix the nuclear 

 changes which occur almost Instantly. We may then study the condi- 

 tions presented by different stages in dead preparations which have 

 been i)roi)erly stained so as to develop the appearance of the nucleus, 

 as little of a trustworthy nature can be learned from any of the twenty 

 species of living eggs which the writer has seen, for in almost all cases 

 the nuclei of living fish eggs are not visible under the microscope, even 

 though magnifying powers of two hundred and fifty diameters be ap- 

 plied. With the use of reagents the matter is much simplified, the 

 nuclei at once become distinct, and their metamorphoses may be very 

 distinctly shown under a power of sixty to seventy-five diameters. 



In Fig. 1 the germinal disk three hours and twenty-three minutes 

 after impi-egnation has been segmented into eight cells ; at the end of 

 four hours and forty-five miiuites it has been segmented into sixteen 

 cells, as shown in Fig. 2. The disk during this time has not increased 

 in transverse diameter, and is relatively smaller, when compared with 

 the vitellus, than the germinal disk of the salmon. It is very transpa- 

 rent, and is less diifereut from the vitellus in color and optical proper- 

 ties than the disk of any fish egg known to rae. It is this feature which 

 makes it hard to find in the live egg, and when found difficult to study, 

 unless the light is skillfully managed so as to bring out the contours of 

 its component cells. The nuclei are still quite invisible in the latter 

 while alive. The stages which immediately follow are still more difli- 

 cult to study, because as the disk spreads to form the blastoderm it 

 becomes relatively thinner and more inconspicuous than in any other 

 lorm known to me, so that it is necessary to manipulate the light in the 

 microscope with extreme caution. 



