472 PROCEEDINGS OF THE NATIONAL MUSEUM. vol. xxix. 



year Ransom reported a similar bud formation in the unimpre^nated 

 eggs of tlie pike. These " showed a lobulation of the concentrated 

 formative .yolk, a sort of irregular asymmetrical cleavage.'' After 

 twentj^-tive hours ''portions of the discus proligerus were pinched ofl' 

 and appeared as projecting buds." His reported in 1899 that unfertil- 

 ized salmon and trout eggs after lying in water four weeks formed 

 hillocks on the surface of the germinal disk by the outpushing of fluid 

 drops under the surface membrane. Neither he nor Ransom give 

 figures. Fig. 29, Plate VII, makes clear these various observations. 

 As to the further fate of the blastodisc in the unimpregnated Qgg 

 of the pipefish, I can only say that it flattens out and finally disap- 

 pears. Fig. 30, Plate VII, is a central section through a blastodisc 

 twent3--six and one-half hours old, which shows this flattening. Fig. 

 31 on the same plate shows a blastodisc taken from a lot of eggs in 

 the invagination stage (forty to fortj^-eight hours). It is much larger 

 and its lower surface is comparatively free from yolk. The contrast 

 is evidently due to the fact that one egg has been lying* free in the sea 

 water, while the other has been under more favorable conditions in the 

 marsupium. Just here it may be of interest to note that while unim- 

 pregnated eggs are often met with in the pouch with embryos of all 

 stages, none of them ever "go bad." Ransom (1866) reports that he 

 has kept unfertilized trout eggs alive in running water forty-three 

 days. More recently. His (1899) gives four weeks for the maximum 

 time, and describes the mass of germ-plasm in the unfertilized eggs 

 of the trout and salmon as decreasing day by day and becoming more 

 and more set through with oil drops and ,yolk spheres. The degener- 

 ating blastodiscs of the pipefish in some cases show these inclusions, 

 but in general ai'e quite free from them. 



YI. SEGMENTATION. 



Before going into a description and discussion of the segmentation 

 of the Q,gg of Siplioatoma floridm.^ I wish to say that this is extraordi- 

 narily irregular. These irregularities ])egin as early as the two-celled 

 stage and become very marked when eight cells are formed. The egg 

 under consideration equals and perhaps exceeds that of the Salmon 

 family in abnormality of cell division. The surface views were nearly 

 all drawn from the hardened germs in 80 per cent alcohol or xylol, 

 the opaque egg making it impossible to draw in situ blastoderms 

 beyond the eight-celled stage. The drawings were all made with a 

 Bausch and Lomb microscope (the tube drawn out to 160 mm) and 

 camera lucida. The surface views were all made with the 1-inch eye- 

 piece and the two-thirds objective. Sections were drawn with the 

 2-inch eyepiece and the one-sixth objective. Plates V and VI have 

 been reduced one-half, the others two-thirds. 



