962 



SCAXDIXAVIAN FISHES. 



wliich show the most niai-ked alterations of growth — 7, 

 8, 10, and 1.'^) in the table — the oldest Fat Herrings 

 from Stavanger (Column 6), though smaller on an aver- 

 age than the Baltic Herrings (Column 2), have advanced 

 further tlian the lattei'. Tlieir development has pro- 

 ceeded more rapidly. Sirailarlj' the youngest Scotch 

 Herrings (Column 7), though inferior in average size 



to the Herrings from Bohuslan (Column 3), have out- 

 stripped the latter in the alterations of growth. Tlie 

 more oceanic its home, the more rapid and more vi- 

 gorous is the development of the Herring type. This 

 is still more clearly shown by comparing with the last 

 two columns Bohuslan Herrings more nearly approach- 

 ing to them in asre: 



The same results may l)c deduced from Lundberg's ' Ocean Herrings with the organs of generation equally, 

 tables by thence selecting for comparison Baltic and or at all events not too unequally developed: 



The Baltic form (the StrOmining) thus fails, even 

 at its most advanced age, to attain averages equal to 

 those shown ))y the much smaller Herrings from Bo- 

 huslan in the last column. The Sti-Ommings, as Nils- 

 son has pointed out, have on an average a larger head 

 and larger ej'es; but this is only a general rule, for 

 exceptions are not uncommon. 



Between the Str5mming and the Herring, however, 

 the averages obtainable from Lundberg's tables show 

 another distinction, the significance of which is all the 

 greater, since it indicates a different direction of deve- 

 lopment or at least a divergent tendency, and seems 

 most naturally to admit of explanation, as we have 

 above stated of the analogous relations in the Salmon- 

 oids, on the assumption either that the forms are still 

 in process of differentiation from each other, or that 



one of them is undergoing a degeneration expressed by 

 reversions to the earlier stages of development. A rule 

 of general validity, which I stated in 1882" in a de- 

 scription of an hermaphroditic example of this species, 

 tells us that, after the dorsal fin has passed through the 

 forward removal which, as Sundevall'' first showed, 

 attends the development of the Herring from the larval 

 stage, there ensues, during the subsequent growth, a 

 backward removal of this fin. In a preceding table 

 (p. 961.) we have also seen that the percentages for 

 the distance between the dorsal fin and the tip of the 

 snout (I'elation no. 5) increase Mith age. The ventral 

 fins share in this backward motion (relation no. 8). 

 Let us now examine these relations in the Stromming, 

 according to Lundberg's tables. For this purpose we 

 divide the specimens measured by hint into four groups, 



" Arch. Biol., vol. Ill, p. 269. It should, however, be reiiiaiked that I was preveuteil from reading the proofs of this paper, which 

 was printed at Brussels, and that a number of misprints, especially in the numerals, have consequently been overlooked. 

 ' Om jiskyngels utveckling, Vet.-Akad. Handl. 1855, p. 20. 



