240 SALMONID^ OF BRITAIN. 



14 rays. Sucli an enumeration, however, is evidently open to uncertainty, unless 

 it is taken into consideration and recorded how many undivided and how many 

 branched rays are present (see fig. 36, p. 163 ante). The first few which are 

 undivided are minute, one or two may even be wanting. Thus in the 8 examples 

 from the same locality already referred to, I found from. 3 to 4 undivided, and 

 9 to 10 divided rays in the dorsal fin, and similarly in the anal, 3 to 4 undivided 

 and 8 to 9 divided rays. Moreau, in France, finds D. -^.^% A. -flf. This 

 character, therefore, is not reliable. 



As to the comparative height of the body, that varies with sex, season, food, 

 and condition, and cannot be otherwise than a very uncertain element in the 

 discrimination of a species. 



Although the difference in the size of the scales has been held as one of the 

 most constant and important characters in salmonoids, one cannot resist the 

 conclusion that such a belief is founded on error. Possibly in no family of fishes 

 are greater variations perceptible in the relative size and number of scales among 

 individuals. 



In examining numerous examples of Salmo fontinalis reared in this country, 

 and counting the pierced scales along the lateral-line, I have found specimens 

 which have from 122 to 142 : but it is in the number of irregular rows which 

 descend from the back to the lateral-line where the greatest diversity occurs, 

 while it is here some naturalists count. Although in 8 of my British char 

 the difference in the number of pierced scales along the lateral-line does not 

 exceed 20, I find from 185 to 235 rows descending to the lateral-line, or a 

 variation of 50 scales. Taking the whole of the reputed British species, the 

 extreme variation recorded is 70 rows. If, therefore, among 8 examples of 

 American char received from two localities I can observe a difference in number 

 of at least 50 rows of scales descending from the back to the lateral-line, the 

 extreme variation of 70 among specimens of char obtained throughout the extent 

 of Great Britain and Ireland, appears hardly sufficient ground for instituting 

 distinct specimens. Sir W. Jardine considered that " the northern, or Suther- 

 land char," has more elongated scales than the "southern char," the scales 

 of which he described as being more orbicular, but having obtained and examined 

 some I do not find such to be the rule. 



" The char of Hawes Water," observed Dr. Davy, " which is known to feed a 

 good deal on insects, is a small and slender fish in comparison with the char of 

 Windermere, which feeds more at the bottom and has a less precarious supply, 

 especially of SquillcB, which abound in the lake.* The one takes the artificial fly 

 freely ; the other — that of Windermere — is rarely so tempted and seldom caught, 

 except by trolling with the minnow. In short," he remarked, " so various are 

 they that in no two lakes do they perfectly agree, either in their average size, 

 form, colouring, or even in their habits." In examining 8 examples from the 

 above locality I found the following variations : — D. -J^, A. flf, L. 1. 118-128, 

 Caec. pyl. (in three examples) 28-42. The number of rows of scales descending 

 to the lateral-line varied from 166 to 180. 



Names. — Oharre, cTiarr or char : torgooh, "red belly" (Wales). Gally-troul 

 "red bellied trout" (Windermere, Lochleven, &c.), while it is also termed at the 

 latter place gelly-trocJi trout (Lochleven Angler), which is said to signify a black 

 leech which abounds in this loch and upon which it feeds. Bed-wame, Scotland. 

 Tarr-dhearg and dearg-bhlian (Gaelic in the Highlands of Scotland). Murneen in 

 Galway and the loughs of Mayo in Ireland (Wilde). L'ombre chevalier, French. 



In olden times the credulous were instructed that three sons of the Church 

 introduced these fishes into Wales from Rome, and placed two in each of the 

 lakes of Llanberris, Llynumba, and Trevennyn. While those in Windermere and 



other the " gilt char," the former keeping chiefly to the upper part of the lake above Bowness. 

 Continuing, " Mr. Braithwaite, of Kendal, who has recently published a treatise on the Salmonidse 

 of Westmoreland, has informed me that he believes in only one species existing in Windermere 

 Lake" (p. 13). 



* Finding the stomach of the char figured in plate ix rather hard, I opened it, but only 

 found two stones which measured as follows in inches, O-3x0-3x0-2 and 0-2x0-2x0-l, 



