44 Transactions. 



'■ erosion or absorption by the living tissue which is known to envelop the 

 scale." Two possible explanations are suggested : " The process may be 

 an anticipatory reduction of the size of the scale to meet the approachin» 

 reduction in the girth of the body, or it may be connected directly with 

 the formation and development of the ova." 



In Canterbury the spawning-mark is by no means so uncommon in 

 trout as it appears to be in England and Norway. With male fish of con- 

 siderable size (say, over 24 in.) it is rather the exception to find scales that 

 do not show a definite spawning-mark — at any rate, in the Selwyn River 

 (see Plate III, fig. 2). In females the act of spawning seems to leave a less 

 decided scar, and most of the cases come within the region of uncertainty 

 mentioned by Masterman, and introduce the personal element. In handling 

 a large number of spawning fish this year, whilst collecting scales, I found 

 that I could in almost every case detect the males by the texture of the 

 skin. The males had a thick tough outer skin, and great difficulty was 

 experienced in removing the scales, whilst no such covering was present 

 in the females, and the scales were easily removed. Under the microscope 

 the scales themselves were in many cases readily distinguished, those of 

 the males being very much more eroded than those of the females. The 

 ripe testes form a very much smaller proportion of the total weight of a male 

 than the ripe ova of a female, so it is natural to suppose that the wastage 

 of tissue in producing the former would be less than in producing the latter, 

 and the shrinkage in milting is certainly less than in spawning, yet the 

 scale-erosion is greater in males. All this seems to suggest that scale- 

 erosion at spawning-time, in trout at any rate, is intimately connected 

 with the production of the thick tough skin assumed by the males. Dahl 

 has noticed that the erosion of scales in spawning salmon is more pronounced 

 in the males, but apparently attaches no significance to his observation. 

 In many cases it is a matter of opinion whether there is a spawning-mark 

 corresponding to any particular winter on a trout-scale, but of the thirteen 

 tagged fish from which I have scales every one shows, if not a distinct 

 spawning-mark, at least a sharply defined winter band, such as the third 

 winter band in Plate I, fig. 1, corresponding to the winter when the fish 

 was stripped and tagged. I think it is probable that such winter bands 

 are tolerably reliable evidence of spawning, but there is an almost perfect 

 gradation from the broad ill-defined bands of the first two winters in 

 Plate I, fig. 1, and many cases must always remain doubtful. 



Dahl (1) assumed that the scales grew in the same projiortion as the 

 fish, and consequently that the distances from the centre of growth to the 

 successive winter bands would be m the same ratio as the lengths attained 

 by the fish in each successive winter. This assumption was almost in the 

 nature of a corollary from what was previously known of the formation 

 of winter bands, but experimental proof was desirable. Dahl and others 

 have collected such a wealth of indirect evidence in favour of this hypo- 

 thesis that there is little danger in accepting it as the basis of my 

 investigations. Direct evidence, however, is difiicult to obtain, and is- 

 meagre. As the whole of the present investigation depends on the truth 

 of Dahl's hypothesis, it will be as well to add my small quota, more 

 especially as Masterman and others have raised the objection that direct 

 evidence is almost, if not entirely, lacking. 



The North Canterbury Acclimatization Society annually strips a number 

 of trout in the Selwyn River for piscicultural purposes, and takes the 

 opportunity to tag two or three hundred fish each year with a small silver 



