Thus the sudden drop in the catch at Sidi-Daud around 1895, which 

 Roule attributes to heavy rains and to the fresh water carried by the 

 outlet of the lake of Bizertep situated 50 km to the west^ finds a more 

 plausible explanation in the analogy of habit with Sardinia, 



e) With the method of the analysis of modes worked out by Vercellij 

 (Atti Congresso Soco It., Progro Scienze^ 1928) there are evident, for 

 some tuna fisheries, fluctuations which are periodical or at least of 

 almost constant period, of various magnitudeso Thus the curve of the 

 catch of the two Sardinian fisheries of Po Scuso and Po Paglia^ taken 

 together, turns out to be composed of periodic elements of 2^ 3-4, 6, 

 10, 24, and 100 (?) yearso The relative calculations for this series were 

 made by Profo Vercelli himself o 



For Sidi-Daud also, it was possible to show fluctuations of quite 

 similar periodo 



Naturally the study of the cases, in part local, in part more or 

 less general, is still all to be doneo 



Granted that the curve of the catch of certain tuna fisheries may 

 be entirely broken up into periodic or almost periodic elements, it_ is 

 possible to make predictions for some of them . 



It is a pity that these statistics, with few exceptions, are very 

 badly kept. The fishery with tuna traps, because of the uniformity of the 

 means (fishing season, position and dimensions of the constituent nets) 

 would lend itself better than any other to analyses of this natureo 



Meanwhile, in a general way, it is possible to give assurances to 

 the operators of the tuna fisheries thax the crisis which has vexed the 

 fisheries of the Mediterranean for some years is certainly transitoryo 

 Apart from fluctuations of lesser magnitude, we are now at the bottom of 

 a cycle of a century (or at least a very long period), and once that is 

 passed over such unfavorable conditions for the fishery will probably 

 not return for several score yearso 



12) The result which is perhaps most interesting in connection with 

 the method of the hooks, is that which leads to the recognition of the 

 existence of a vast area frequented by tuna of identical raceo It is 

 this which permits one to attempt an overall outline of the movements 

 which the tuna make in connection with reproduction and feedingc 



This area embraces the Mediterranean, about half of the North 

 Atlantic as far as the Azores j, and the North Sea, from about the latitude 

 of the Tropic or a little farther north to the Arctic Circle at the 

 other extreme o 



The tuna is not very common at the Canaries (where the other purely 

 Atlantic or more tropical species of tuna predominate) and certainly it 

 does not push much farther south along the ccast of Africao To the north 

 it is common in the region of Trondhjem in Norway and it has been seen 

 frequently at Lofoten, 



14 



