Mr. E. \N'. L. Holt on the Ova o/Gobius. 37 



picro- carmine tlie pedicle of attachment in our ova takes the 

 carmine stain very deeply, whilst the zona (and external 

 membrane it' present) is quite unatiected by it. 



This seems to induce the belief that tlie process of attach- 

 ment is similar in nature to filaments of an external mem- 

 brane rather than representing an everted membrane as in 

 Osmerits. 



Turning to the question of the parent fish, Professor 

 31'Intosli has kindly given me a drawing of the eggs of 

 Gohius niyer from the Ciianncl Islands, which will be seen to 

 present some points of resemblance to those under discussion. 

 Both are considerably elongated and both possess filamentous 

 processes at their lower ends. In G. nujer^ however, the 

 meshwork is less distinct tiian in ours and the t^Z^ are fixed 

 in rows by the interlacing of their filaments. In both forms 

 the perivitelline space would aj)pcar to be large *. 



Ot the nature of the yolk in Professor Prince's specimens 

 or in Gohius niger I am unable to speak. The yolk in our 

 eggs, however, presents great likeness to that of a larval 

 form common in this bay and long since identified by Pro- 

 fessor M'Intosh with a species of goby. Two species are 

 common here, viz. G. Ruthensparri and G. minutus. In the 

 Seventh Annual Keportof the Scotch Fishery Board Professor 

 M'Intosh, writing " Un the Pelagic Fauna of the Bay of 

 St. Andrews during the months of 1888," mentions young 

 gobies (chiefiy G. minutus) as occurring in some numbers in 

 the net in July and August, ranging in size from o'o to 11 

 niillira. Their occurrence much earlier (at stages too young 

 for diagnosis of species according to our present knowledge 

 of this genus) is frequent, but their ova have never been 

 found here, probably because other forms have occupied 

 attention. A few larval forms appeared this year in April 

 and May, of one of which I append a figure (fig. 6). Larval 

 gobies are readily distinguished by the characteristic pigmen- 

 tation and very early appearance of a conspicuous air-bladder. 

 The specimen figured measured 3*57 raillim. The anus is 

 slightly anterior to median, the pectorals are large and fan- 



* Professor Prince, writing to me from Valentia under date Mav 25, 

 1890, describes some eggs that had just been found, during the cruise 

 under the auspices of the Royal Dublin Society, in the pools about the 

 Beginnish Islands. In shape they strongly resemble those of G. nu/er, 

 and they are placed side by side on end, as in our form. From a rougli 

 sketch the perivitelline space appears very large ; the length is about 

 -jV inch. Professor Prince is developing these eggs, and will no doubt 

 be able to throw some light on to their species. Judging from the tact 

 that a female G. Ruthensparri was taken in the same pool, he thinks it 

 probable that they belong to that species. 



