338 THE REPRODUCTION OF TELEOSTEAN FISHES 



hardly any perforations near the micropylar region, the numerous 

 recticulo-radiate ridges being mostly united by a thin membrane. 



G. scorpioidcs is certainly the smallest British Goby so far recorded. 

 If common it is not likely to be often retained in the meshes of an 

 ordinary net. 



Aphia pellucida. Nardo. 



We cannot find a description of the ova of this fish, though in other 

 respects, thanks to Collett,* the cycle of its life-history is fairly well 

 known. In the early part of July of the present year, adults of both 

 sexes were rather numerous on the zostera and weed beds of the inner 

 part of Cawsand Bay, the females being full of roe. They became 

 scarce towards the end of the month, and none have been since taken. 

 The abundance of this species in the estuary of the Lynher in April 

 has already been noticed by one of us. {Jonrn. M. B. A., vol. i., p. 89.) 



None could be found there on the 21st July of the present year. 

 If the fish is a permanent inhabitant of the estuary this would seem 

 to indicate that the brood of last year, ofi'spring of the half-grown 

 examples met with in April, had already died off, having fulfilled 

 the life-span of a single year allotted to tliem by Collett. | It is, 

 however, possible that individuals move seawards from the estuary 

 at the approach of maturity, while the larvae in turn migrate to the 

 estuary. 



Though breeding adults were numerous, we failed to find the spawn 

 attached to any object trawled or dredged in Cawsand Bay. Examination 

 of the ovary of a female, 1| inches long, taken in the Bay on the 14th 

 July, indicates that the ova are certainly demersal. The yolk, probably 

 not quite mature, was transparent, colourless, and almost free from 

 granulations. It consisted at this stage of an outer layer, enclosing 

 an inner and more refractive part. After 15 hours in sea- water the 

 ovum was evidently dead. It had become opaque and yellowish, the 

 refractive part having been apparently broken up into a number of 

 globules, bearing a general resemblance to those of Gobius. Under 

 natural conditions it is probable that the yolk is not essentially dif- 

 ferent from that of Gobius, but leys opaque and without conspicuous 

 colouration. The zona is thin and without special markings. As in 

 Gobius, it is enveloped in the ovarian condition by an outer membrane, 

 which is everted when the follicle is ruptured, and forms the fixing 

 apparatus. This outer membrane is, however, divided into a number of 



• Proc. Zool. Soc, 1878, i., p. 318. 



t The brood of 1898, if present, would he loo small to lie letaiueil in the net enipkiyed. 

 It is iiossible that some larvie, taken at the mouth ol' the Lyiiher in 1897, may have 

 belouged to these species ; tluy \\eic not preserved. 



