852 



SCAXDIX AVIAN l-'ISHi:: 



tlie height of the iuiul fin. Uiit in liie young this elia- 

 racter loses its validity, as well as in very old male 

 Salmon during the spawning-season. Tlie length of 

 tlie maxillaries is, as a rule, less in tlie Salmon, greater 

 in the Trout, than 8 % of that of the body or 3(5 % 

 of that of tlie head; hut the ciianicter, whieh was ori- 

 ginally sexual, ajiplies only to typieal specimens. Sal,- 

 mon, as mentioned above, in general have comparatively 

 larger scales than Trout. Glnther and Day have ex- 

 ])ressed this b^- the statement that, in an oblique trans- 

 verse row tbrward from the posterior extremity of the 



Fig. 212. Vomer of a Salmon {A) aiirl a Tronl (B), seen from 

 below (f/) ami from the left {h). Natural size. After v. Siebolp. 



adipose fin to tlie lateral line, the Salmon has 11 or 

 12 scales, the Trout 13 — 15. lUit even tliis charai'ter 

 holds good only in tj-pical specimens", and sometimes 

 the two sides of the body differ in this respect'. 



Among the internal characters that may be used 

 to distinguish Trout from Salmon, Ave have al)ove 

 mentioned the number of tlie gill-rakers, which in the 

 outer ro^v on the front of the lirst bi'anchial arch is, 

 as a rule, under 17 in the Trcnit, over 17 in the Sal- 

 mon, or, if it be greater in the former, made up in 

 front and at the anterior upper corner of verrucose, 

 spiniferous protuberances. Another character, derived 

 from the shape and dentition of the vomer, has receiNcd 



special attention of v. Siiciioi.n. In tvjtical Salmon the 

 head of the vomer has a inure or less distinct pent- 

 agonal or hexagonal shape, in Trout it is triangular. 

 In the former the vomerine teeth disappear earlier, and 

 in old Salmon are sometimes almost entirely wanting: 

 whereas even the oldest Trout commonly have several 

 teeth not only in a transverse row on the liase of the 

 head of the vomer, but also in a double or single row- 

 along the carina on its shaft. 



But these internal characters are no more trust- 

 worthy than the external: the naturalist \vho examines 

 hundreds of specimens will find them too often merged 

 in each other. 



i^ole. One of tlie best examples to show the relation of the 

 two varieties to each other, as regards the last-mentioned characters, 

 seems to be afforded by the attempts to introduce the typical Salmon 

 into the Lake of Geneva''. These experiments were commenced by 

 flie hatching of Salmon cffgs from the Rhine and from the hatchery 

 of Hiiningen; but the fry excluded from these ova escaped into 

 the Rlione, and none of tliem was ever regained and identified with 

 certainty. From 1857 to 1860, according to Chavannes, about 7,000 

 fry were placed in the feeders of the lake, and during 1863 about 

 4,600; and in 1882 a pisciculturist, LUGRIN by name, is said to 

 have deposited nearly 100,000 Salmon fry in the Rhone near Geneva. 

 Hardly any results, according to both Lunel and Fatio, are known 

 to have been attained by these experiments: the 'Salmon' which liave 

 been caught from time to time in the Lake of Geneva, have proved 

 on closer examination to be Trout. The latter author, however, is 

 still dubious with regard to at least three finds. In these specimens 

 Trout and Salmon characters were blended, and especially in two of 

 them the vomer was most nearly approximated to its Salmon type. 

 Though it is indeed possible that all the Salmon fry have been de- 

 stroyed, it seems more probable that during growth the fry have 

 passed to the Trout form, as this appears in its normal development 

 where the water offers to the fish no practicable means of commu- 

 nication with the sea. 



The Salmon is so well known that a minute de- 

 scription would be here superfluous, especiall}- as it 

 differs but slightly from the Charr, which we have 

 just described, ^^'e may instead refer the reader to 

 our figures and to the following comparative table of 

 tlie mo.st important external characters, expressed in 

 averages which have been computed 1) in the first 

 two columns, from our measurements of Trout in the 

 Forell, Ori)ir/, and Salmon stages and of t}-piral Sal- 

 mon at the same periods of growth^ 2) in the third 

 column fi'om all these measurements combined for each 

 character, and 3) in the fourth column from the cor- 

 responding measurements of Scandinavian Cliarr. 



triitUi _riin-uifilis (the Freshwater Trout) Day assigns in his last work {Brit., h: SiibnonuUr, p. 109) 12 — 15 scales 



RihsHiuseets Salmonidev. 



" To Sal 

 in the said row. 



'' See, for example, specimens Kos. 284, 285 and 306 in Tabula nictrica W in Smitt' 

 ' See LuNEL, 1. c, pp. 128 — 130, and Fatio, 1. c, pp. 308—313. 



