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



season in which they were captured than by stating merely 

 the historical order of their addition to the collection. More- 

 over, the data for any one year were not sufficient in them- 

 selves to elucidate the problem of the rate of growth of the 

 species. The calculated age, owing to the fact that young 

 gobies may be hatched at any period from the middle of 

 April to the end of August, can only be a rather crude 

 approximation to the truth. This method of calculation 

 rested primarily on the average rate of increase in length per 

 month. It seemed probable that by taking for the first 

 example a comparatively old specimen, there would be less 

 risk of error in obtaining this average. Accordingly the 

 average length of six specimens captured in St. Andrews Bay 

 on 5/9/89 was taken ; this was nearly 75 mm. Now the 

 greatest number of larval forms occurs, as has been stated, in 

 June; so that it may be inferred that the time when the 

 maximum number of these is hatched is the end of May and 

 beginning of June. Combining these conclusions with the 

 fact that larval gobies are about 1*5 mm. in length on 

 hatching, it may be stated, generally speaking, that a speci- 

 men of G. minutus hatched at the beginning of June will 

 increase in length 75 — 1*5 = 73'5 mm. by the beginning of 

 September of the following year, i. e. in fifteen months' time. 

 From this the average rate of increase in length per month 

 may be deduced, viz. 7 -||=4'85 mm. It seemed more than 

 probable that the rate of growth would be accelerated during 

 the spring and summer and retarded during the late autumn 

 and winter. The otoliths of the specimens averaging 75 mm. 

 were examined, and occasion will be found to deal with these 

 more minutely later on. At present it is only necessary to 

 remark that the otoliths showed a thin zone, which had every 

 appearance of indicating a retarded period of growth. One 

 of the largest specimens of O. minutus captured in St. An- 

 drews Bay during its first summer measured 20 mm. : the 

 date of capture was Sept. 17th. This fish would be rather 

 over 3^ months old if it had been hatched at the most prolific 

 spawning-period. During that time it had increased 20 — 1*5 = 

 18*5 mm.; the average rate of growth per month is -§qf = 

 5*30 mm. It will be noticed that an exceptionally large 

 first-year form was taken during September and measuring 

 30 mm. This fish must have been one of the earliest of the 

 season ; its age is estimated at 4| months, but even this 

 estimate gives an increase of G'OO mm. per month. However, 

 the rate of growth of no fish is so even that it can be ex- 



