232 Mr. E. Shann on the Life-history and 



rings on its scales. Applying this argument to the subject 

 of the present paper, it may be supposed that during- the 

 whole of their short life the gobies are more or less active, 

 that they continue to grow uniformly, and correlative with 

 this we tind the even growth of their scales. It would be 

 necessary to keep some gobies under observation during a 

 complete year in order to prove the above hypothesis. When 

 the otoliths are examined it will be found that there is in 

 them an indication of the periods of growth characteristic of 

 the better known food-fishes. Thus, there may be other 

 causes at work which bring about the evenness of the con- 

 centric rings on the scales. The radial markings on scales 

 have also been suggested as possible means for age determi- 

 nation. It has been observed by Cunningham *, and is 

 evident in the present drawings from the scales of different 

 species of Gobius (see PI. IX. figs. 4-7), that the radial mark- 

 ings can be divided by their respective lengths (on the same 

 scale) into two or more groups. The first group consists of 

 those markings which reach, or nearly reach, the innermost 

 concentric ring of the scale. The second group of markings 

 terminates approximately at the concentric ring which formed 

 the limit of the first season's growth. In larger fishes more 

 groups are found ; but, as might be expected, the gobies never 

 show more than two series, indicating two periods of growth. 

 It is in the cases of G. niger and G. Jeffrey sii that this 

 arrangement of the radial markings is most accentuated ; the 

 markings on the scales of the 51 mm. specimen of G. minutus 

 (PI. IX. fig. 4) offer little evidence in this respect, and from 

 the knowledge of this species it may be stated that the radial 

 markings in all cases which have been examined offer as little 

 help as the concentrics towards age determination. 



In all cases the scales were taken from the side of the fish 

 described, halfway between the anterior ends of the soft dorsal 

 and anal fins, unless otherwise stated. 



(1) G. niger. 



Scales comparatively large, rarely more than 40 in a longi- 

 tudinal row, 13-lo in vertical f, irregular square in outline, 

 strong radial markings, teeth long compared to those on the 

 scales of other gobies. A specimen taken in July and 



* XXIII. Annual Report S. F. B. p. 125. 



f These numbers, and those given for other species, are taken from 

 Holt and Byrne, op. cit. The numbers have been verified where the 

 individuals of any given species in the collection were in sufficient 

 quantities. 



