GoDBT. — Growth of Brouui Trout in Canterbury. 49 



more marked in the fourth year. A close examination of the figures in 

 Tables I (A) to I (F) will show that in practically every case there is one 

 year for each fish in which it has made more than normal growth ; it may 

 be the second, third, fourth, or fifth year, but in almost every case there 

 is this break in the growth-curve. This sudden jump or break is generally 

 attributed to migration to more favourable surroundings, and there is every 

 reason to believe that this is the case with the Selwyn fish. Practically 

 all the spawning in the Selwyn takes place in the shallow, shingly part. 

 Except in the spawning season, fish of any considerable size are rare in this 

 part of the river. The traps are set just about the junction of the shallow 

 water and the deep, to catch the fish working up to the spawning-beds. 

 Consequently every fish caught has come from the deep water. Probably 

 every fish was hatched and spent its early youth in the shallow part of the 

 river ; therefore at some period it must have migrated to deep water. An 

 examination of its scales will generally disclose when that migration took 

 place The average curves, therefore, are really compounded of a number 

 of different curves representing one-, two-, three-, four-, and possibly five- 

 and six-year-old migrants. In fig. 3 are shown typical curves for a two- 

 year-old and a four-year-old migrant. Plate III, figs. 1 and 2, show 

 scales from these fish respectively, in which the period of better growth 

 subsequent to migration is very distinctly shown. Whenever an average 

 growth-curve closely approximates to a straight line for four or five 

 years it is a fairly definite indication that the fish from that locality are 

 migratory. 



Table I (E) shows the figures for thirteen trout caught last summer 

 with rod and line at the mouth of the Selwyn and other streams running 

 into Lake Ellesmere. The average rate of growth is about the same as 

 that of the 1915 males, or intermediate between those of the 1918 males 

 and females. 



In order to ascertain whether results in any way reliable could be 

 obtained from smaller samples I calculated the average growth for the 

 first, second, third, &c., twenty fish in Table I (B). Considering the very 

 complex nature of the water, the agreement is quite satisfactory, and 

 indicates that results. of some value can be obtained from quite small 

 samples. 



Trout in the Selwyn, whatever the mode of growth, seem to have a more 

 or less fixed limit of growth at about 23 in., which is rarely exceeded. 

 Other waters also seem to show a maximum size-limit. It is curious, 

 however, that this limit is occasionally considerably exceeded, and not 

 necessarily by very old fish. These abnormally large fish, so far as I can 

 ascertain, show no peculiarity of growth common to all, but their scales 

 seem on the average unusually broad in proportion to their length, though 

 I am at present unable to state this definitely. Whether the large size is 

 determined by heredity or by unusually favourable environment I cannot 

 say, though I am inclined to attribute it to the former. It is certainly 

 a point worthy of further investigation. Particulars of five of these 

 abnormally large fish are given in Table I (F). 



Rivers. 

 From the angling point of view the rivers of Canterbury may be 

 divided into two classes — snow rivers and rain rivers. The former contain 

 large trout, for the most part, of sea-going habits ; the latter comparatively 

 small trout, which are not as a general rule migratory. 



