BIHANG TILL K. SV. VET.-AKAD. HANDL. BAND 27. AFD. IV. K:0 4. 17 



It is thus necessary to find another proof of tlie signifi- 

 cance of the averages, and the most instructive that I have 

 fovmd is derived from the relation between the height of the 

 body at the beginning of the anal fin and the distance of 



this fin from the tip of the lower jaw \ " %\. There it 



will be seen (table V), that the forms named palUdus and 

 seminudus always take lower figures t han Vahlii and reticii- 

 latus respectively, at the corresponding age (as far as age 

 can be judged from the length of the body); and for semi- 

 nudus indeed in this relation there seems to be a constant 

 character, distingnishing it from reticulatus, as its highest 

 limit of variation is below 22 %, what in reticulatus occurs 

 only in the larval state (a < 60 mm). This character, how- 

 ever, seems to be due to a divergency of the manner of evolu- 

 tion or, perhaps rather, to its interruption for seminudus, as 

 will be seen from the following averages: 



a, va. mm. 

 altul „ 



Here for 'pallidus and Vahlii the fignres evidently are 

 increasing with age, for reticulatus this increment is rather 

 small (if a real one), but for seminudus the fignres are a 

 little decreasing from the yonnger to the older age, perhaps 

 the beginning of a specific divergence, if it not at last, with 

 a fuUer material for examination, should turn ont to be more 

 properly regarded as the retention of a juvenile character, 

 in which case seminudus in this respect would hold the same 

 position to reticulatus, as does frigidus in other respects. 



A few words may be added about a character, to which 

 many authors have given a great weight. It is well known 

 that in the Vahlii-grow^ as well as in the frigidus-form the 

 most conspicuous lateral line is the ventral one. while in the 

 reticulatus-gToii^ the median lateral line appears in this cha- 

 racter. But besides, in the former group usnally at least 



