344 TRANSACTIONS LIVERPOOL BIOLOGICAL SOCIETY. 



deeper gulley at D, which descends to over 50 faths., and a 

 shallower area at S. If we suppose that, as is actually the 

 case in many localities, such Copepoda as C alarms helgo- 

 landicus or Euchceta norvegica are at certain times of the 

 year confined to the deep water below 50 faths., then they 

 would occupy the shaded area at D which is below about 

 two-fifths of the surface of the water, and in volume is 

 about one-fifth of the whole. If four equidistant vertical 

 hauls (1, 2, 3 and 4) are taken across the loch, these 

 Copepoda (say, C alarms) will appear in one quarter of the 

 samples, and if vertical hauls 1 and 2 only are taken 

 Calanus will appear in one half of the material, while if 

 hauls 2, 3 and 4 are taken no samples of Calanus will be 



Pig. 19. Diagram showing diverse vertical hauls. 



obtained. In every one of these cases, if the haul or hauls 

 were taken as representative of the whole area, the result 

 obtained would be erroneous. The real distribution in 

 this case is that Calanus occupies one-fifth of the whole 

 mass of the water ; but if one considered only the number 

 of square metres underneath which vertical hauls showed 

 Calanus, the conclusion arrived at would be two-fifths, 

 twice too much. If haul 1 only were examined, it might 

 be thought that Calanus was universally present, if haul 3 

 that it was absent, and if hauls 1 and 2 that it was present 

 in half the water, much too large a result — the conclusion 

 being that results obtained from few and distant hauls 

 are liable to be deceptive. 



