io6 THE DRY-FLY MAN'S HANDBOOK 



man being invisible until the fish is nearly tired out 

 and practically in the net. Weeding is in many cases 

 the result of a scare on the part of a trout from seeing 

 the fisherman or feeling the strain of the rod. 



The Test trout generally are in good fettle during 



the first half of September, and 

 September. on the Itchen, which, as before 



noted, is a later river than the 

 Test, they are in good condition to the end of the 

 month. I doubt whether the policy of killing Test 

 trout after the middle of the month is a wise one. 

 Some years back I suspected that the majority of the 

 fish killed were then females getting heavy with roe, 

 and since then continual examination has convinced 

 me of the good grounds for this suspicion. The 

 result of continuing the season after, say, September 

 15th must therefore be to unduly diminish the stock 

 of female fish likely to spawn at the end of the 

 year. 



Mr. Sheringham suggests that it might be of use 



to anglers to give them a non- 

 Non-scientific deter- scientific means of differenti- 

 mination of sex of . , ^ , 



atmg the sexes 01 the trout. 



With the view of obtaining 

 the best possible opinion on the subject, I consulted 

 Mr. G. A. Boulenger, of the British Museum (Natural 

 History), Cromwell Road. He wrote me as follows : 

 " 1 am afraid it would be impossible to draw up 

 a definition of the two sexes in Salmo fario which 

 could serve in all cases without reference to the 



