44 
bulletin oe the bureau of fisheries. 
In order to carry out tests in which the objection could not be raised that the results 
might be due to the shock of cutting nerves rather than to the loss of a sense organ, the 
following procedure was employed: By taking two stitches of very fine silk thread, 
one on either side of the external olfactory aperture, it was comparatively easy to close 
this aperture, and thus to prevent any passage of water through the olfactory sacs. 
Killifish which previous to the operation gave markedly different and characteristic 
reactions to the two classes of packets already described reacted to both kinds of packets 
after their anterior olfactory apertures were closed, as they had previously done to the 
packets that contained no food. That this reaction was not to be directly attributed 
to the operation of stitching up the apertures was demonstrated in two ways. If, after 
the stitches were taken, the thread was not drawn up and tied so as to close the aper- 
ture, but the ends were allowed to remain free, the fish would react as normal fishes 
do to the two classes of cloth packets, thus showing that the mechanical injury due to 
the stitches themselves did not influence the fish in any essential way. Further, if fishes 
whose anterior olfactory apertures had been closed by stitching and tying, and whose 
discrimination for the two classes of packets had thereby been lost, had their olfactory 
apertures reopened by cutting and removing the thread, they very soon regained their 
capacity to distinguish packets with food from those without food; in other words, they 
soon returned to the condition of normal fishes. For these reasons it is believed that 
stitching up of the anterior olfactory aperture is in itself not a disturbing operation for 
the fish and that the loss of the ability to recognize the presence of hidden food is in 
reality due to the loss of the. olfactory function. 
CONCLUSIONS. 
Whether such olfactory reactions as those that have just been described are really 
due to smell or not is regarded by some authors as an open question. Nagel (1894, p. 
56), who has discussed this matter at some length, concluded on rather theoretic grounds 
that fishes could not possibly possess a sense of smell and that their so-called olfactory 
organs act more as organs of taste than of smell. Possibly the whole matter is merely one 
of definition. With human beings smell differs from taste chiefly in the concentration 
of the stimulating solution and not, as was formerly supposed, on the state of the stimu- 
lating material, for, though we usually say that we smell gaseous or vaporous materials 
and taste liquids and solids, all these substances are in reality dissolved on the moist 
surfaces of whichever sense organ they stimulate. The most striking difference between 
smell and taste with us is that we smell extremely dilute solutions and taste only very 
much more concentrated ones. As a result we recognize the presence of many distant 
bodies by smell and not by taste, for the very minute amount of material that reaches 
us from the distant body will form a solution on our moist surfaces that will be stimu- 
lating for our organs of smell, but not for our organs of taste. Hence our olfactory 
organs as compared with our organs of taste are what Sherrington (1906) has called dis- 
tance receptors, a designation justly emphasized by Herrick (1908). Although this 
distinction between taste and smell is one of degree rather than of kind, it seems to us 
