550 JOURNAL, BOMBAY NATURAL HISTORY SOCIETY, Vol. VIII, 
food for the consumption of animals—all animals either living directly on 
vegetable food, or being dependent on it indirectly, as in the case 
of the carnivora which eat the herbivora. Animals in turn furnish 
carbonic acid to the plants for food, and when they die, are ultimately 
converted into salts and gases which may serve as food for planis. 
Thus we get a cycle, a constant succession, of changes of life ending 
in death, and of death ending in life. But in the deep sea this series seems to 
want a link, for we have no plants to manufacture food out of inorganic sub- 
stances. The wanting link, however, is only apparently wanting. Whence 
then the ultimate food supply? It is supposed that it all comes from the 
surface. ‘The surface of the sea is crowded with organisms, both animal and 
plant, and there appears to be an incessant rain of small food from the surface. 
Great rivers also carry into the sea a large amount of water-logged matter that 
finds its way to the bottom. On food derived from these sources, the smaller 
and lowlier forms feed, and the larger and higher forms feed in turn on the 
lowlier forms. This necessity for catching prey impresses a very distinct 
mark upon many deep sea animals, which are horribly rapacious. This is best 
seen in certain fishes which have the teeth enormously magnified, and the 
mouth capable of even greater distention than that of a serpent. In this spe- 
cimen of Thaumastomias, the lower jaw is practically a great rat-trap attached 
to the head simply by a long elastic muscular band, and capable of being 
turned completely over the head. In Ponerodon we have a somewhat similar 
mouth with large hinged teeth : from the stomach of this specimen of Pone- 
rodon I removed the remains of a fish weighing more than the Ponerodon itself, 
the animal being a mere appendage of its own abdomen. In Polycheles, again, 
observe the long chelipeds, and that all the “ feet” end in chele. 
The third factor we may now consider is, that which covers the modifica- 
tions due to the action of the law, “ avoid being killed and eaten.” “ Protec- 
tive resemblances” for the purpose of deceiving enemies appear to be rare in 
deep sea animals, and we can easily understand that it must be so, seeing that 
there is no light. Coloration, which is so largely developed in land animals 
for purposes of protective resemblance, cannot come within the scope of 
natural selection where light is wanting. Hence deep sea animals are com- 
monly sombrely and uniformly coloured in harmony with the uniform sombre 
gloom of their habitat, deep black being common, and deep purple shading 
off to red and pink. 
The action of the fourth factor—sexual selection—appears also to be much 
hampered by the want of light. However, our knowledge of the deep sea 
fauna is not yet comprehensive enough to allow us to make any definite 
statements on this subject. So far the only case of difference in form between 
the male and female which has come to my notice, is in this :specimen :of a 
male of Neobythites pterotus, which has the pectoral fin-rays much prolonged, 
perhaps for the purpose of feeling for the female. At any rate, in the female 
the rays are of the ordinary length. 
