THE BLENNY FAMILY. 211 



opaline, yolk-spherules. Outside the roe is a transparent mem- 

 brane, not readily removed from the surface. 



In the recent work on Scandinavian Fishes^ the authors had 

 overlooked previous remarks on the subject, so that they only 

 observe that " the ovaries of the female are full, and the milt- 

 sacs of the male distended at the end of October," thus showing 

 that the spawning-season occurs in that month, or soon after. 

 As Malm, however, found one of 31-5 mm. in July, and others 

 are recorded by Kroyer, the spawning-season, Malm thinks, 

 must be later. 



The free eggs of the gunnel were first noticed by Mr 

 Anderson Smith, and they adhere to each other like those of 

 the herring or sea-scorpion. 



They were next observed at St Andrews between tide-marks 

 in masses about the size of a Brazil-nut in holes of the boring 

 shell-fish, Pholas, at the Pier rocks, and almost every season 

 they have been found between December and March in similar 

 positions. Moreover, several examples have deposited their 

 eggs in the tanks of the Laboratory, so that Mr Holt was 

 enabled to draw up an account of the oviposition and early 

 stages. When observed in the holes of Pholas the parent-fishes 

 were coiled beside the eggs, and the same happened when they 

 spawned in the tanks of the laboratory. Those in Shetland 

 seem to be somewhat later in spawning, since eggs at the 

 hatching point occurred in April. Mr Holt found that on 

 the 10th February one had spawned and that the female was 

 engaged rolling up the mass of fertilized eggs into a ball (Plate 

 VIII, fig. 6). The body was so curved that the head rested on 

 the tail, the mass of eggs being held in the loop thus formed. 

 Occasionally the body was constricted so tightly that the eggs 

 slid over the back, again to be encircled. Sometimes the 

 male, which is the smaller, alternately encircled the eggs 

 with the female. "The object is evidently to prevent the ova, 

 which are adhesive only when first extruded, from being 

 scattered and lost, and it may be supposed that the operation 

 is performed before the eggs are deposited in the narrow 

 crevices in which they have been found." This care of the eggs 

 1 Fries, Ekstrom and Sundevall, 2nd edit. (Smitt), 1893, p. 223. 



14—2 



