THE SENSE-OEGANS AND PERCEPTION OF EISHES. 237 



The process of searcli is equally indirect and tentative by day and 

 by nighty wlietlier tbe food is exposed or bidden in an opaque 

 vessel, wbetber a piece of actual food is in tbe water or tbe juice 

 only, squeezed tbrougb a cloth, and, lastly, wbetber (as tested in 

 tbe case of. tbe conger and rockling) tbe fisb be blind or not. On 

 tbe otber band, if tbe olfactory epitbelium is destroyed in tbe rock- 

 ling or conger, tbe animal does not leave its biding-place to bunt, 

 tbougb it seizes food placed near its face. Stones or otber objects 

 rubbed witb food attract tbese fisbes as mucb as food itself, and 

 wben very bungry tbey will snap at sucb uneatable substances, 

 tbougb tbey are rejected after being taken into tbe moutb. I saw 

 no reason to suppose tbat any of tbese animals bave tbe instinct of 

 following a scent (as a prawn seems to do to some extent), tbougb 

 tbey always stop to examine bodies upon wbicb food bas lately 

 rested. Tbe scent of tbe food does not seem to remain long in tbe 

 water, and apparently tbe scent of tbe surface of tbe food itself is 

 dissipated or decomposed in a sbort time ; for wben, for instance, a 

 piece of squid is not found after searcbing, it often bappens tbat tbe 

 fisbes give up and retire, but will come out again in quest if tbe 

 same piece of squid be taken out, cut in balf to expose a new sur- 

 face, and put back. None of the fisbes were ever seen to bunt for 

 more than about fifteen minutes unless tbe scent was renewed. It 

 is difficult to estimate tbe distance to which a scent can be diffused 

 in tbe water, but it is likely that, in water which is not rapidly 

 moving, its virtues are destroyed before it bas been carried far. 

 There is, I think, no reason for supposing that scents are diffused 

 tbrougb the water otherwise than by currents. This is most easily 

 tested by experimenting witb shrimps in a large shallow vessel. 

 The shrimps remain buried until tbe scent reaches them. Wben 

 tbe water was not in motion, if food was gently dropped in, the 

 shrimps gave no sign for an indefinite time, but on stirring the water 

 tbey began to seek. The longest tank at Plymouth is about twenty 

 feet long, and an interval of from five to ten minutes elapses before 

 conger at one end are aware of the presence of food put in at tbe 

 other. 



The perceptions, then, by which these animals recognise tbe pre- 

 sence of food are clearly obtained by means of tbe olfactory organs, 

 and apparently exclusively through them. I was particularly sur- 

 prised to find no indication of the possession of sucb a function by 

 the sense-organs of tbe barbels and lips, or by those of tbe lateral 

 line. As bas been already described, tbe pelvic fins and barbels of 

 tbe recklings {Motellci) and tbe lips, &c., of most fisbes bear great 

 numbers of sense-organs closely comparable in structure witb the 

 taste-buds of otber vertebrates. No one who has seen the mode of 



NEW SERIES. — VOL. I, NO. III. 18 



