There is, in other respects than those mentioned, much more diversity than in most other families, although the species are not very numerous; their exact number, however, is uncertain. The species appear to be subject to unusual variation, and characters which would be of specific importance in most groups in this appear to be merely individual deviations. A result of this was that before the extent of this variability was appreciated many specific names were based on what later proved to be unstable characters. For instance, an able French naturalist, Prof. Émile Blanchard, thought that in France alone there were seven species with three dorsal spines and four species with many. He was even strongly inclined to think that other species (plusieurs espèces) remained to be discovered.

Naturalists of the present day reduce the seven to one (Gasterosteus aculeatus) and the four to another (Pygosteus or Pungitius pungitius). It is not impossible, however (but not probable) that it may be found that from the extreme of undue multiplication or "splitting" of species there has been a revulsion to an opposite extreme of excessive "lumping." Much more certain are the genera: these are readily separated and stand out boldly in nature. Three of these occur in the Old World-Gasterosteus, Pungitius (Pygosteus), and Spinachia. Gasterosteus and Pungitius are equally well represented in North America, and the absence of Spinachia is more than balanced by the existence here of two unknown to Europe-Eucalia and Apeltes. Of all these the nesting habits are more or less known.

All the sticklebacks take care of their eggs and the newly born young, but it is the male, and not the female, that exercises parental care; he it is that builds a nest that would do credit to a bird and drives or entices the full female to enter into it and deposit her ripe burden. When a sufficient supply of eggs has been secured, the male closes the nest and remains in charge till the young have reached a size which he considers to be sufficient to enable them to wander away and seek their own living.

The sticklebacks have been divided among three groups or subfamilies.

FIG. 80.-Ten-spined Stickleback (Pungitius pungitius).

After Storer.

FIG. 81.-Gasterosteus.
Pelvis below.

The best-known forms, or Gasterosteines, have the pelvic bones forming an arrow-like or V-shaped plate below, the point extending backward.

The three-spined sticklebacks (Gasterosteus), ten or many spined sticklebacks (Pygosteus), and those with four or five dorsal spines reclinable directly backward (Eucalia) belong to it.

One peculiar American type, the Apeltines, have the pelvic bones

FIG. 82.-Four-spined Stickleback (Apeltes
quadracus). After Storer.

FIG. 83.-Apeltes. Pel-
vis below.

obtuse behind, lateral bars, and between them an unarmed belly. A third type, the Spinachiines, have the pelvic bones loosely connected in front only and the ventral fins inserted farther back than in the others; the body is also more elongated, the snout extends forward in a somewhat tubiform manner, and the dorsal spines are

more numerous.

The most generally distributed and characteristic of the genera and that to which the largest species except Spinachia belong is Gasterosteus, or the three-spined sticklebacks. The body is comparatively high and has a fusiform contour, there are two large free spines on the back and a smaller one in front of the fin, and the ventral spines are considerably behind the roots of the pectorals and under the interspace between the first and second dorsal spines. The most common species-common to Europe and America-is the Gasterosteus aculeatus.

Almost any permanent body of water, however small, may harbor some. In the words of Smitt, "It is often met with in collections of water so small and so isolated that it appears difficult, if not impossible, to explain the manner in which it has been conveyed thither, or the sources from which it derives its support. It is fondest of calm water, and in summer frequents shallow spots close inshore, especially where the sunshine has full play. Here it leads a merry life, and one may often see it leap several centimeters out of the water, while at other times it keeps still at the same spot, as though there were nothing in the world to disturb it. But in a moment it is all life and spirit and darts off with the speed of an arrow. In stormy weather it is tossed on the waves and has thus been cast even into a boat. Toward autumn it retires to deeper spots on the shores of channels, estuaries, or larger inlets. In late autumn and at the beginning of winter, however, it roams about in large companies." It then prepares for winter quarters; "when the cold is most severe, it probably lies packed in large shoals at the bottom."

In

It is an eminently gregarious fish and its schools are sometimes very large, but they are not equally large every year. Some years they may be comparatively few and in others exceptionally abundant. many of the northern streams, W. C. Kendall (1902) testifies, they occur "in such vast numbers as to be used for fertilizer" and "as food for cows and dogs, and even for men. They were taken in large numbers in the brush weirs used for catching small herring on the coast of Maine, and in the same locality often become a nuisance by clogging the nets of the swelt seiners."

One instance illustrative of their occasional extraordinary numbers has been often quoted, but is as apt now as ever. In 1776 Pennant claimed that "once in seven or eight years amazing shoals appear in the Welland" Canal "and come up the river in the form of a vast column. They are supposed to be the multitudes which have been washed out of the fens by the floods of several years and collected in some deep hole till, overcharged with numbers, they are periodically obliged to attempt a change of place. The quantity is so great that they are used to manure the land, and trials have been made to get oil from them. A notion may be had of this vast shoal by saying that a man employed by the farmer to take them has got for a considerable time 4 shillings a day by selling them at a half penny a bushel."

This account has been quoted as unparalleled, but several notices in American publications come nearly, if not fully, up to it. In the Canadian Annual Report on Fisheries for 1863 (p. 61) it is reported that the three-spined stickleback or picassou was "caught in great quantities in the small rivers, brooks, and barachois of Magdalen Islands, where it is used as food for cattle and as manure," and that "400 barrels were caught" in 1862 "in the barachois of Basque Harbor " alone. Four hundred barrels were also caught in 1866 and sold as manure at 25 cents per barrel, but in 1867 the catch was smaller (150 barrels) and prices advanced to "1s. 3d. per barrel.” «

Sticklebacks are very voracious and almost omnivorous, their rapacity being only limited by their size. The eggs and fry of other fishes suffer severely from their attacks, but with apparently equal relish they take worms, the minute entomostracan crustaceans, the larvæ and imagines of insects, and small mollusks. But they do not refrain from attacking fishes much larger than themselves. "Mr. Mable, at the Weston-super-Mare Museum, had some three-spined sticklebacks in an aquarium, and some roach, Leuciscus rutilus, were added. With this invasion the prior inhabitants were dissatisfied, but not frightened, as they forthwith attacked the newcomers, biting at them anywhere until they became thoroughly cowed. These little tyrants were observed to place themselves in front of the roach,

a Fisheries Appendices for 1866 (p. 42) and 1867 (p. 34).

steady themselves by their tails, and then suddenly dart straight at the lips of their intended prey, from which they bit pieces out. These attacks were continued until all the roach had been killed, when they were eaten by their conquerors."

Another instance of insatiable voracity has been recorded by Baker. One devoured in five hours 74 young dace, which were a quarter of an inch long and of the thickness of horsehair. Two days afterwards it swallowed 62, and would probably have eaten as many every day could they have been procured for it.

Such are the main facts respecting the general habits of the sticklebacks, but their chief interest to us are the extraordinary nests which they make and the still more extraordinary provision with which nature has endowed the males to make those nests.

When spring and warm weather are well advanced males and females become impressed with the instinct to reproduce their kind, and the whole organization is modified accordingly. The female's abdomen becomes distended with eggs, and the male, who had been silvery white during winter, assumes a brilliant livery, develops from the urinary bladder and kidneys the material for a peculiar threadlike substance, and prepares to make a nest for the reception of eggs of such females as he may be able to influence.

The peculiar secretion from the kidneys deserves some notice here, for it is unique and only known in the sticklebacks. It was first fully described by Karl Mobius (1885) and soon afterwards by John Ryder (1882) and E. E. Prince (1885). The last has not only given an excellent account of the peculiar structure, but some apt illustrations, and these are reproduced herewith. (Plate I.)

EXPLANATION OF PLATE I.

After E. E. PRINCE.

Fig. 1. Transverse section (male Gasterosteus sp.) of renal mass, urinary bladder, etc., in situ, × 150 diameter. A, kidneys (coalesced portion): a, a, right and left ureters passing longitudinally along outer ventral borders of kidneys; b, urinary tubules, muclei of epithelial cells indicated. B, urinary bladder, behind the cervix: a, epithelial; b, fibrous layer; c, c, right and left ureters (on opposite sides of bladder); d, contained thread-like secretion from kidneys; C, intestine; D, D', testes.

Fig. 2. Fragment of hyaline capsule of ovum, showing the rows of pits and lamellæ, X 300.

Fig. 3. Fragment of hyaline capsule of ovum, in transverse section, showing numerous lamellæ, X 350.

Fig. 4. Portion of mucous secretion, showing funicular structure, × 350. Fig. 5. Sketch of nest (diagrammatic), one-third natural size, the pockets containing ova exposed. a, a, transeversely arranged intersecting threads; b, b, masses of ova contained in the interspaces of nest.

Fig. 6. Dissection of male, showing viscera of posterior portion of abdominal cavity, about natural size. a, enlarged urinary bladder; b, left flexure of urinary bladder (cervix); c, right flexure of urinary bladder (near posterior termination); d, genital pore; e, cloacal depression; f, anus; g, urinary aperture; 1. alimentary canal; i, i, right and left testes; j, j, kidneys; j', ureters; k, swim bladder.