OCCURRING IN THE NEIGHBOURHOOD OP PLYMOUTH. 37 



The character of the ovum is shown in Fig, 25. Its diameter, mea- 

 sured in a direction parallel to the surface of attachment, is 1"2 mm. 

 The embryo was distinctly formed and somewhat advanced in 

 development, the heart having begun to beat ; but this organ lying 

 on the surface of the yolk anterior to the embryo is not shown in 

 the figure. The yolk is of an orange-red colour and made up of 

 separate minute yolk-spherules ; it also contains near the tail of the 

 embryo a number of oil-globules of different sizes. The tissues of 

 the embryo were very transparent. The chromatophores are limited 

 in number, intensely black in colour, and confined to the dorsal 

 portion of the yolk-sac near the tail of the embryo. 



Blennms galerita, and BlerDnus pJiolis are stated by Day to have 

 adhesive ova, the former depositing them on stones, the latter in 

 holes in rocks, on the authority of Couch (Zoologist, 1846). Centro- 

 notus gunnellus has adhesive ova which adhere together and form 

 a free round mass. Zoarces viviparus hatches its ova in its ovary 

 and produces about fifty young at a time all alive and similar to the 

 parent except in size (see my paper in Trans. Roy. Soc. Edin., 

 188G), Shore fishes, like fresh-water, fishes have usually adhesive 

 ova, or in some cases heavy ova which sink to the bottom. Such 

 ova usually have oil-globules, either a single one or several, and the 

 yolk is always made up of minute yolk-spherules. 



Callionymus lyea (the Dragonet). 



I have already referred to the ova of this species in connection 

 with my observations on Solea vulgaris, and have stated that I 

 identified the ova taken by the tow-net shown in figs. 26 and 27 as 

 the ova of Callionymus lyra. I also mentioned that Raffaele denies 

 altogether the existence of the hexagonal marking of the vitelline 

 membrane in the Mediterranean species G. fesUvus. On the other 

 hand, Eaffaele found an exactly similar reticular marking of hexa- 

 gonal meshes on the vitelline membrane of the fertilized ova of 

 Uranoscopus scaher, and in the ovarian ova of Sanriis lacerta. It is 

 certainly somewhat inconsistent on the part of the Italian zoologist 

 that he should assume that Mcintosh mistook an epithelium for a 

 marking of the vitelline membrane in G. lyra, and should affirm the 

 existence of the marking in Saurus on evidence exactly equivalent 

 to that on which Mcintosh relied. Saurus is a genus of the 

 Scopelidae, which family belongs to the order Physostomi. 



Leaving the question of Saurus lacerta entirely aside for the 

 present, the similarity between the ova of C. lyra and Uranoscopus 



