I 



INVESTIGATIONS IN FORTH, 1889 AND 1890. 141 



dabs was below (by 3) that of 1888. Cod and gurnards re- 

 mained about the same in each year, as did others. The 

 increase in the number of the "witches" in this and the 

 previous year is a fact of interest. 



All the Stations except IV., YI. and IX. showed a reduction 

 in the average of the food -fishes captured. Station IV. had 

 only an increase of 6, yet this was the highest average (272) 

 of the year. The increase per haul at Station VI. was 31, and 

 at Station IX. only 8. 



In contrasting the diminished averages of 1889 with the 

 previous year it must not be forgotten that the former has the 

 disadvantage of including five winter months in its working- 

 period, whereas the latter has but two. Under these circum- 

 stances it is perhaps remarkable that the reduction was not 

 greater. Yet such does not indicate that a serious diminution 

 of the fishes of the area had taken place ; but is the effect of a 

 well-known seasonal variation. 



A considerable increase took place in the captures of 1890, 

 seventy-nine hauls producing a total of 18,270 fishes or an 

 average of 231 per haul. Of these 14,953 or an average of 189 

 per haul were saleable, and 3317 or an average of 42 unsale- 

 able. The work giving this result was carried on during 10 

 months, five of which were winter months, viz. February, March, 

 April, November and December, and five warmer or productive 

 months, viz. May, July, August, September and October. While 

 the three first mentioned winter months, however, produced 

 respectively 501, 720 and 891 fishes of all kinds, the two last 

 months of the year gave respectively for seven hauls 1970 and 

 for nine hauls 1753 fishes, a very considerable addition to the 

 totals. The captures in the warmer months were as follow : — 

 May with 9 hauls 1479, July with 7 hauls 2432, August with 

 6 hauls 2284, September with 5 hauls 2067, and October with 

 9 hauls 4173. It is clear that if July, August and September 

 had had as many hauls as October the result would have been 

 very different, for these are productive months. In an inquiry 

 like the present, therefore, it is important to have the data 

 equivalent throughout, or at any rate to be cautious in drawing 

 deductions from results obtained under different conditions. 



