8 071 the Emhryonic History o/" Pteroplattea micrura. 



obliteration of the communication between the umbilical canal 

 and the intestine^ become a placenta. 



Now from the non- vascular condition of the yolk-sac and 

 umbilical canal in the embryos under consideration we may 

 venture to surmise the possibly equally direct origin of 

 aplacental viviparity from simple oviparity. 



Here ag'ain the large egg remains in the terminal portion 

 of the oviduct, and in the process of development external 

 gills which had originally " very possibly become specially 

 developed to facilitate respiration within tlie egg " * become 

 otherwise specialized to absorb nutriment from a yolk-sac 

 which has only the single communication with the embryo 

 through the umbilical canal and intestine. When the yolk is 

 all finished the nutriment which is secreted from the maternal 

 glands naturally follows, in the absence of any absorptive 

 blood-vessels in the empty yolk-sac, the already established 

 route through the branchial clefts, one of which (the spix-acle), 

 being unobstructed by gill-filameuts, becomes at last the 

 exclusive channel of supply. 



I should like, in concluding this paper, to express my obli- 

 gations to Professor Howes for his extreme kindness in sending 

 out to me on loan, at great risk owing to distance, his own 

 copies of Herr Jaekel's and others' papers on the subject of 

 the affinities of the Batoids — an act of kindness and considera- 

 tion which a ship's naturalist, cut off for months from all but 

 a few standard classics, can hardly over-appreciate. 



EXPLANATION OF PLATE IV. 



Fiy. 1, Embryo of Fteroplatcea micrura, from dorso-laleral aspect ; nat. 

 size, but with only a few of the gill-filameiits represented, for 

 the sake of clearness, s, spiracle. 



Fig. 2. End of a gill-filament, showing the marginal capillary filled in 

 places with blood-clot. X 42. 



Fiff. 3. Transverse section of a gill-filament, showing the marginal capil- 

 lary in section and the single fold of epithelium. X 188. For 

 the sake of clearness the blood-clot is represented in one limb of 

 the capillary only, and the spaces between the nuclei of the 

 surface epithelium are a little exaggerated. 



Fig. 4. End of a trophonema, or nursing-filament, seen as a transparent 

 object in glycerine, showing the marginal artery and the super- 

 ficial capillary plexus. X 42. The mediau vein is not seen so 

 near the end. 



Fig. 5. Obliquely transverse section through a nursing-filament, showing 

 the glands still in the form of solid bulbs lying beneath a still 

 unbroken surface of epithelium. X 110. a, a, arteries ; v, 

 vein ; c, c, superficial capillaries. 



* Balfour, up. et torn. cit. p. 62. 



