867 



Changes of growth in 

 the true Salmons. 



Leiljilli uf 111.- Iu';ul r.-,liicLHl in "i »i Ihi- Ini-I 



Base of the duvsal lin „ „ „ „ „ 



Length of the pecloral lins „ „ „ „ „ 



PostabiJoininal length t, <« i, .* t> 



Height of the anal tin „ „ „ „ 



Least (lcl)lh of the tail , ., ., 



Length of the middle caudal rays 



Length of the maxillaries in % of the length c,f the 



Breadth ., „ „ „ „ „ „ „ „ ,, 



h Mt' llie h.»lv 



head re.l 

 mnxilliir; 



Us 



.s.o 



14.3 



11.3 



IS.o 



ir),3 

 i-.'.i 



lti.3 

 19.9 

 l-.>.3 

 8.8 

 7.9 



o8.i 



15.4 



•2(!.7 



14.9 



20.9 



r)'j.4 



12.4 



21.0 



Nil iimiiediate descent of the Snliiinns from the 

 Smelts rail tliu.s be assumed. 



<)ii examining the relations l)et\veeii tlie three 

 forms liitherto distin;i-uisiied in tiie genus of the Smelts, 

 we tinil tliat tlie Atlantic Smelt {Osmerus epcrlantts) 

 (listiiietl\- a|)|)ears in some respects to have adopted its 

 own course of development, and that in these respects 

 it has remained nearer to tlie jiresumable oriijin of the 

 genus. But in most points the development seems to 

 have been common to the three forms, and the Pacific 

 Smelt {OsiiL (Iciifex) is apparently the nearest repre- 

 sentative of the original type in the modern fauna. 

 We are besides struck by the analogy in the relations 

 of development to those between the three above-men- 

 tioned Charr forms, Salmo alpinus, S. stagnalis, and 

 S. salcelinus, for we tind witliin the genus of the Smelts 

 too a defined natural series of three forms, 



Osmerus deiitex — 0. spirinchus — 0. eperlaiius, 

 as the most general expression of the course of evt)lu- 

 tion. Several deviations, however, meet us, which may 

 here too be explained as manifestations of the influence 

 of sex on the development, the most Arctic form — 



Osincriis splrinchus\ known from Alaska, Siberia, and 

 the White Sea — being differentiated in most respects 

 from the other two forms by the predominant in- 

 fluence of the female characters, while Osinenis cpcr- 

 lanus, generally speaking, bears the stamp of the male 

 characters. 



Ill order to find an expression for the general 

 alterations of growth within the genus, we have dis- 

 tributed our specimens among four classes according 

 to age, three of these classes fixed by comparatively 

 wide gaps in the decrease or increase of the percen- 

 tages during growth, the fourth (the oldest) marked !))• 

 a reversion to the earliest characters, best explained, 

 perhaps, as the result of a growth exceeding the most 

 natural limits of the forms. In certain parts of Swe- 

 den a distinction is made in popular language' between 

 n<ir>< (the smaller Smelts) and slum (the larger ones), 

 and at a length of 200 mm. there intervenes in the 

 Scandinavian Smelt one of the said gaps, for example 

 in the decrease of the percentage for the relative di- 

 mensions of the eyes. At a length of about l.jQ mm. 

 we observe a similar li-ai), and at a length of aljout 



" Cf. Smitt, Riks 



Sab) 



p. '.). 



p. 12. 



' 1. c, p. 12. 

 '' 1. 



p. 14. 

 ' I. L-., p. 14. 



/ 1. ... p. ir,. 



'J 1. I-., p. 16. 



* 1. c, p. 21. 

 ' 1. c, p. 21. 

 J 1. c, p. 23. 



* The name of spirincliKs was applied by LiLUEBoitr, (Vet.-.\kad. Handl. 1850, p. 304) to the White Sea Smelt, and although Pallas, 

 who first coined the name, by no means restricted it to that form, we have refrained from giving the said form a new name, as spirinclms 

 in its old signification cannot be employed for the fnture as a specific name. In our first classification of the Salraonoids in the Koyal 

 Museum (Ofvers. Vet.-Akad. Fiirh. 1882, No. 8, p. 34 and Gt. Intern. Fish. Exhib. Lond. 1883, Swed. Cat., p. 184), we called this species 

 Osmerus dvinensis. 



' See Arteui, Descr. Spec. Pise, p. 47 and Xilsson, Shaiid. Fiia, Fish., p. 433. 



