600 or 800 individuals. When young they have no fixed habitation, but encamp sometimes in one place, and sometimes in another, under the shelter of their web; but when they have attained two-thirds of their growth, they weave for themselves a common tent. About sunset the regiment leaves its quarters. At their head is a chief, by whose movements their procession is regulated. When he stops all stop, and proceed when he proceeds; three or four of his immediate followers succeed in the same line, the head of the second touching the tail of the first; then comes an equal series of pairs, next of threes, and so on, as far as fifteen or twenty. The whole procession moves regularly on with an even pace, each file treading in the steps of those that precede it. If the leader, arriving at a particular point, pursues a different direction, all march to that point before they turn.1

The following additional facts concerning these remarkable habits may be quoted. I take them from the account published by Mr. Davis in 'Loudoun's Magazine of Natural History: '

The caterpillars, he observed, were Bombyces, and were seen crossing a road in single file, each so close to its predecessor that the line was quite continuous, 'moving like a living cord.' The number of caterpillars was 154, and the length of the line 27 feet. When Mr. Davis removed one from the line the caterpillar immediately in front suddenly stood still, then the next, and next, and so on to the leader. Similarly, those behind the point of interruption successively halted. After a pause of a few moments, the first caterpillar behind the break in the line endeavoured to fill up the vacant space, and so recover contact or communication, which after a time it succeeded in doing, when the information that the line was again closed was passed forward in some way from caterpillar to caterpillar till it reached the leader, when the whole line was again put in motion. The individual which had been abstracted remained rolled up and motionless; but on being placed near the moving column it immediately unrolled, and made every attempt to get readmitted into the procession. After many endeavours it succeeded, the one below falling into the rear of the interloper. On repeating the experiment by removing a caterpillar fifty from the head of the procession, Mr. Davis found that it took just thirty seconds by his watch for information of the fact to reach the leader. All the same results followed as in the previous

1 Kirby and Spence, Entomology, Letter xvi.

case. It was observable that the animals were guided neither by sight nor smell while endeavouring to close up the interrupted line; for the caterpillar next behind the interruption, on whom the duty of closing up devolved, 'turned right and left, and often in a wrong direction, when within half an inch of the one immediately before him; when he at last touched the object of his search, the fact was communicated again by signal; and in thirty seconds the whole line was in rapid march.' This gentleman adds that the object of the march was the search for new pasture. The caterpillars feed on the Eucalyptus, and when they have completely stripped one tree of its leaves, they all congregate on the trunk, and proceed as described to another tree.

De Villiers1 gives an account of his observations on the manner in which these caterpillars (Cnethocampii pitzocampa) are able to pass information, which does not quite agree with the above observation of Mr. Davis. For he says that, in a train of 600 caterpillars, interference by him in any part of the train was communicated through the whole series instantaneously-all the 600 caterpillars stopping immediately and with one consent like a single organism.

In

According to Kirby and Spence there is a kind of caterpillar (Pieris crategi) which lives in little colonies of ten or twelve in common chambers lined with silk. one part they make of the same material a little bag or pocket, which is used by the community or household as a water-closet. When full of excrement the caterpillars empty it by turning out the pellets with their feet.2

Only two other instances of noteworthy intelligence as exhibited by larvæ have fallen within my reading. One of these is mentioned by Réaumur, who says that the larvæ of Hemerobius chrysops chase aphides, and having killed them, clothe themselves in their skins; and the other case is the very remarkable one mentioned in his newly published work by W. MacLachlan, F.R.S., of caddis-worms adjusting the specific gravity of their tubes to suit that of the water in which they live, by attaching heavy or light material to them according as they require sinking or flotation.

1 Trans. Ent. Soc. France, vol. i., p. 201.
2 Introduction to Entomology, Letter xxvi.

241

CHAPTER VIII.

FISH.

ALTHOUGH We here pass into the sub-kingdom of animals the intelligence of which immeasurably surpasses that of the other sub-kingdoms, it is remarkable that these lowest representatives of the higher group are psychologically inferior to some of the higher members of the lower groups. Neither in its instincts nor in general intelligence can any fish be compared with an ant or a bee-a fact which shows how slightly a psychological classification of animals depends upon zoological affinity, or even morphological organisation. For although a highly competent authority, namely Van Baer, has said that a bee is as highly organised an animal as a fish, though on a different type,' no one would be found to assert that an ant or a bee is so much more highly organised than a fish as its higher intelligence would require, supposing degrees of intelligence to stand in necessary relation to degree of organic development. And this consideration is not materially altered if, instead of regarding the whole organism, we look to the nervous system alone. There is no doubt that the cerebral hemispheres of a fish, although small as compared with these organs in the higher Vertebrata, are, bulk for bulk, enormous as compared with the oesophageal ganglia or 'brain' of an insect; while the disproportion becomes still greater if the cerebral hemispheres of a fish are compared with their supposed analogues in the brain of an ant, viz., the pedunculated and convoluted lobes which surmount the cephalic ganglion. But here the relative smallness of the ant as a whole must be taken into con1 Phil. Frags., translated by Huxley, Taylor's Mag., 1853, p. 196.

R

sideration, and also the fact that its brain is relatively much more massive as well as more highly organised than that which occurs in any other order of invertebrated animals, except, perhaps, the octopus and his allies. Therefore, although the brain of a fish is formed upon a type which by increase of size and complexity is destined in function far to eclipse all other types of nerve-centre, we have to observe that in its lowest stage of evolution as presented to science in the fishes, this type is functionally inferior to the invertebrate type, where this reaches its highest stage of evolution in the Hymenoptera.

Emotions.

Fish display emotions of fear, pugnacity; social, sexual, and parental feelings; anger, jealousy, play, and curiosity. So far the class of emotions is the same as that with which we have met in ants, and corresponds with that which is distinctive of the psychology of a child about four months old. I have not, however, any evidence of sympathy, which would be required to make the list of emotions identical; but sympathy may nevertheless be present.

Fear and pugnacity are too apparent in fish to require special proof. The social or gregarious feelings are strongly shown by the numberless species which swim in shoals, the sexual feelings are proved by courtships, and the parental by those species which build nests and guard their young. Schneider saw several species of fish at the Naples Aquarium protecting their eggs. In one case the male mounted guard over a rock where the eggs were deposited, and swam with open mouth against intruders. The following accounts of the nidification of certain species of fish show that the parental instincts are not unlike those which obtain in birds, and are comparable in point of strength with the same instincts as they occur in ants, bees, and spiders.

Agassiz remarks 1 that while examining the marine products of the Sargasso Sea, Mr. Mansfield picked up and brought to him a round mass of sargassum, about the size of the two

1 Silliman's American Journal, Feb. 1872.

fists placed together. The whole consisted, to all appearance, of nothing but gulf-weed, the branches and leaves of which were, however, evidently knit together, and not merely balled into a roundish mass. The elastic threads which held the gulf-weed together were beaded at intervals, sometimes two or three beads being close together, or a branch of them hanging from the cluster of threads. This nest was full of eggs scattered throughout the mass, and not placed together in a cavity. It was evidently the work of the Chironectes. This rocking fish-cradle is carried along as an undying arbour, affording at the same time protection and afterwards food for its living freight. It is suggested that the fish must have used their peculiar pectoral fins when constructing this elaborate nest.

The well-known tinker or ten-spined stickleback (Gasterosteus pungitius) is one of our indigenous fish which constructs a nest. On May 1, 1864, a male was placed in a well-established aquarium of moderate size, to which, after three days, two ripe females were added. Their presence at once roused him into activity, and he soon began to build a nest of bits of dirt and dead fibre, and of growing confervoid filaments, upon a jutting point of rock among some interlacing branches of Myriophyllum spicatum-all the time, however, frequently interrupting his labours to pay his addresses to the females. This was done in most vigorous fashion, he swimming, by a series of little jerks, near and about the female, even pushing against her with open mouth, but usually not biting. After a little coquetting she responds and follows him, swimming just above him as he leads the way to the nest. When there, the male commences to flirt-he seems unaware of its situation, will not swim to the right spot, and the female, after a few ineffectual attempts to find the proper passage into it, turns tail to swim away, but is then viciously pursued by the male. When he first courts the female, if she, not being ready, does not soon respond, he seems quickly to lose his temper, and, attacking her with great apparent fury, drives her to seek shelter in some crevice or dark corner. The coquetting of the male near the nest, which seems due to the fact that he really has not quite finished it, at length terminates by his pushing his head well into the entrance of the nest, while the female closely follows him, placing herself above him, and apparently much excited. As he withdraws she passes into the nest, and pushes quite through it, after a very brief delay, during which she deposits her ova. The male now fertilises the eggs, and drives the female 1 Ransom, Ann. and Mag. Nat. Hist., 1865, xvi., p. 449.