ME. F. DAT 0>* BEITISn SALMONES. 401 



milt yielded 40 per cent, of young fish, but uoue if the milt of 

 the charr were used ; that the ova of a hybrid between a trout 

 and a charr could not be fertilized with trout milt. I saw at 

 Berlin lovely hybrids between trout and charr, but was unable to 

 obtain satisfactory evidence that such were fertile forms. 



I Tvill now briefly record the results Avhich I have arrived at 

 in testing the interesting conclusions, given in Dr. Giinther's 

 ' Catalogue,' as to what constitutes a species of trout. I took 

 first what are termed constant characters, as the number of the 

 vertebrae as well as of the csecal appendages, as they apj)eared 

 to be considered among the most important factors in affording 

 a guide to specific differences. I obtained leave from Mr. Elwes 

 to use his trout-preserves at Colesbourne, on the summit of the 

 Cotteswold Hills, where no new races of trout have been intro- 

 duced, and consequently the original local form remains un- 

 changed. According to Dr. Griinther's investigations, the variety 

 of brook-trout termed S.fario, distributed in the northern parts 

 of Europe and Scotland (Catal. vi. p. 59), has vertebrae 59-60 

 and caeca pylorica 33-46 ; whereas the variety Ausonii is stated to 

 possess vertebrae 57-58 and caeca pylorica 38-47. The latter " is 

 found in Central Europe and the southern parts of England " 

 (vi. p. 59). Certain varieties of distribution are alluded to ; ana 

 it is asserted that the northern form " extends as far southward 

 as Shropshire, where both forms are met -^dth." I captured a 

 considerable number of trout at Colesbourne, which from its 

 locality should have produced the variety Ausonii, and found that 

 they had the number of vertebrae stated to belong to that form, or 

 57-58, but that their caecal appendages were 34-39, or apper- 

 taining to the northern race. This rendered it clear that reliance 

 could not be placed on these figures ; the proposed formula of 

 vertebrae and caecal appendages were not found correlated at 

 Colesbourne on investigation, and therefore could not be depended 

 upon as invariable in other places. The next locality from which 

 the examples came that I minutely investigated Avere from Car- 

 diganshire, already referred to ; and here again an anomalv was 

 found. The number of vertebrae were from 57 to 60, but the 

 caecal appendages from 35 to 44. Tabulated, they would be as 

 follows : — 



Sahnofario, northern form (Giinther), Vert. 59-60, csee. append. 33-46. 

 „ southern „ „ „ 57-58, „ 38-47! 



„ from Gloucestershire .... „ 57-58, „ 34-39 



„ „ Cardiganshire „ 57-60, „ 35-44! 



