FRESH-WATER TROUT— C^CA AND VERTEBRiB OP. 189 



dissecting a rainbow trout, Salmo iridens, at Howietoun, in March, 1887, which 

 had been hatched and reared in that establishment, instead of 41 of these 

 appendages which it is said to possess in its native hill streams, it had 71. 

 And it is common for some continental varieties living in large lakes to have 

 these appendages augmented in number. 



Again from some cause, as will have to be again referred to when discussing 

 the Lochleven trout, when removed to localities where they deteriorate, the 

 number of these appendages may become reduced. 



Instances of variation in the number of these appendages are as frequent as 

 variations in the external colours of the fish, probably even more so, for local 

 causes occasion forms to have some colour characteristic of the locality, whereas 

 in the number of ca3ca no such constancy prevails. In the foregoing instances it 

 may be roughly stated* that, excluding the Lochleven race of trout, the number of 

 appendages observed to the south of the British Isles has been found from 33 

 to 51, and in Scotland from 27 to G9. While the southern form in New Zealand 

 has developed from 33 to Gl. As to the Lochlevens they have been recorded 

 between 48 and 90 : the only constancy in short to be found in these organs 

 consists in their inconstancy, but taking a wide expanse of rivers and waters for 

 examination, they ai'o as a general rule more in number to the north in the 

 British Isles than to the south, similar in fact to what obtains in the races of 

 sea trout. 



It has seemed to me that at the commencement of the angling season I have 

 found in some waters where ti'out subsist largely on water snails, Limnea, that the 

 coats of the stomach are thicker at these periods than they are after the May fly 

 season when they have a more varied diet and do not appear to restrict themselves 

 so much to these snails. 



Vertebrw, respecting the number present in fresh-water trout, I have found 

 from 57 to 59 in Sutherlandshire, from 57 to GO from the Windrush in Glouces- 

 tershire, 57 to 59 from the Churn on the Cotteswold Hills and 5G to 57 in the 

 Tivi. Mr. G. Sim, a.l.S., examined fourteen examples from Aberdeenshire : in 

 five males he found 58, 58, 58, 58, 59; in three from Kincardineshire, 59, 59, 59. 

 In six females from Aberdeenshire 58, 58, 58, 59, 59, 60, showing a variation in 

 numbers from 58 to 60. 



Dr. Giinther found in the northern form from 69 to 60, and the southern form 

 57 to 58; out of the fourteen examples in the British Museum collection there is 

 one from the Lyne with 59, one fronr the Isle of Man with 58, one from Northum- 

 berland GO, three from Cumberland 58, 58, 59, two from Shropshire 58 and 60, 

 two from the Usk 58, 58, one from the Rhymney 58, one from Buckinghamshire 

 58, and two from Hampshire 57, 57. 



The foregoing figures would seem to demonstrate that the number of vertebras 

 in the fresh-water trout, irrespective of whether it is taken to the north or south 

 of the British Isles, varies between 56 and GO. 



It is very evident that there is as close a relationship between the anadromous 

 marine trout and the non-migratory fresh- water forms as between the anadromous 

 and the land-locked salmon. Thus the anadromous sewin or sea trout of 

 Welsh rivers crosses with brook trout, and fertile ofl'spring are the result (see 

 p. 144a«/e). So also does the so-termed Galway sea trout, S. estuarius or cjallivensis: 

 while the phinoc or whitling (p. 160 ante) passes by similar gradations into the 

 brook trout, or as observed by Sir H.Davy in Sahnonia, p. 70, " The river and sea 

 trout seem capable of changing permanently their places of residence." 



I will now pass on to enumerate the various forms of fresh- water trout existing 

 in the British Isles, and which have been termed species. In thus separating 

 them such is done merely for the sake of easy reference to the writings of 

 author.s, but not with any idea that most of them even deserve to be classed as 

 distinct races. 



* Doubtless increased investigations will show far more variations in number than given 

 above. 



