fishes to be selected for the aquarium; collecting together in schools, tame, hardy and lively, they have qualities which few aquarial specimens possess. The stickleback (Gasterosteus) of which there are several varieties, is hardly a fish for the general collection; although of exquisite form, it is so fierce, especially in the breeding season, that it incessantly attacks the other fishes in the aquarium, and in a short time deprives them of more or less of their tails, making the unfortunate victims literally top-heavy, swimming with their tails, or rather what were once tails, much higher than their heads.

Sticklebacks should have a tank devoted exclusively to them and this especially if we wish them to build a nest, one of their peculiar accomplishments. Early in the spring the sticklebacks may be found in great numbers in the small ditches which drain the salt-water marshes. The male is easily distinguished from the female by its deep red color around the gills and its blue eyes, while the female has only the silvery scales. A pair taken at random usually live peaceably together; if it is in the right season they will soon look about for materials for a nest, taking bits of water-plants and even coming to the surface for small pieces of straw and sticks; with such materials they build a round nest about as large as a small English walnut, hollow in the centre and having two holes large enough to admit the fish on either side; the nest is built upon the branches of some of the water-plants. While the female is laying the eggs the male acts as guard, fiercely driving away anything coming within a certain radius of the nest. When the eggs are laid they resemble small globules of wet sago more than anything else. The female will be seen to fan these eggs quite often with her fins; this is probably to give them fresh water and to prevent any sediment from collecting upon them. After a fortnight or so, instead of eggs, we see in different parts of the tank what at first look most like very minute gold spangles as large as the head of a small

pin. On closer examination we find that they are the eyes of a very small fish. Their growth is so slow that in order to preserve them it will be well to remove them to a small tank by themselves, where they can be fed by placing a piece of raw beef on the end of a string, and hanging it over the edge of the tank into the water until it is turned white, when another piece can be introduced. The stickleback, as also the minnows, is easily accustomed to fresh water by freshening the salt-water gradually until it is quite fresh and then introducing the fish into the tank. The stickleback is not the only fresh water, nest building, fish. Wood mentions a curious fish found in tropical America, called by the natives the hassar; a fish which builds a nest as carefully as the stickleback, though one "not placed in the water but in a muddy hole just above the surface." Whether we have gold fish or not in the aquarium, is a matter of taste, some persons thinking that they give the aquarium a common fishglobe look. It seems to me if we can get some small ones of a brilliant color, and of good proportions, we should be glad to receive them into the tank. The great trouble with gold fish is that they are apt to be so deformed, some with the gaunt look of a starved fish, others with a hump on the back or a larger or smaller number of fins than usual. Gold fish would be worth keeping in the aquarium for their remarkable color alone if for nothing more.

Small eels and horned pouts add to the variety of fishes in the aquarium, but both are so uneasy and so very voracious that they are not pleasing inmates of the tank; wandering up and down the sides of the tank, they seem discontented and ill at ease. Young alewives are so beautiful that one is tempted to try them in the aquarium; rarely do they flourish in it.

One of the most interesting animals for the aquarium is the triton, or water-newt; these tritons are often found in what are called, in the country, pond holes, seldom in brooks or ponds; they are perfectly harmless and will remain on the

1

warm hand as long as one has patience to hold them; they come up to the surface to breathe, and therefore do not consume much oxygen; they are perfectly hardy and easy to keep alive, eating small pieces of beef eagerly; they occasionally change their skins, bringing the old skin over their heads and then swallowing them just as toads do. Their odd motions in the water, often poising themselves on the end of the tail or on one toe, are very amusing. They lay their eggs in the early spring either on or between the leaves of water-plants. By the middle of August the young are nearly two inches long; they breathe at first with gills, but by September they come to the surface for air, as the older ones do. These tritons outlive all the other specimens in the tank, and they live so peaceably with their companions that they are invaluable as aquarial specimens.—To be concluded.

REVIEWS.

THE DEVELOPMENT OF INSECTS.-Naturalists are now paying increased attention to the embryology of the articulates. After Rathke, Herold, and Kölliker had published their memorable works, there was an interval of twelve years between the publication, in 1842, of Kölliker's celebrated tract, entitled in Latin, "Observations on the first Genesis of Insects," and Zaddach's "Researches on the Development and Structure of Articulated Animals; Part I. The Development of Phryganidan Eggs,” which appeared in 1854. Then followed Leuckart's "Propagation and Development of the Pupipara, from observations on Melophagus ovinus ;” Huxley's article in the Linnæan Transactions, on the "Reproduction and Morphology of Aphis;" and Lubbock's essay on the "Ova and Pseudova of Insects," in the London Philosophical Transactions for 1859. Claparède, in 1862, published his splendid and beautifully illustrated prize essay on the "Evolution of Spiders," and a year after Weismann followed with a succession of brilliant works on the "Embryology and Anatomy of Diptera," which are in many respects the most important essays on the embryology of the hexapodous insects that have yet appeared, while the illustrations, very copious and detailed, are the most elaborate we have yet seen. Of great importance also is Mecznikow's "Researches on the

Embryology of the Hemiptera (Aphis, Aspidiotus, Corixa), and Simulium, and the viviparous Cecidomyian larva," which were printed in Siebold and Kölliker's Journal, in 1866.

At the meeting of the American Association, in August, 1867, the writer presented a paper on the Development of a Dragon-fly (Diplax), an illustrated abstract of which appeared in the NATURALIST, vol 1, p. 676. His studies did not embrace the earliest changes in the egg, but only those observed after the rudiments of the head and appendages appeared. In 1868 Dr. Alex. Brandt presented to the Imperial Academy of Sciences, of St. Petersburgh, a paper entitled "Contributions to the Developmental History of the Libellulidæ (Calopteryx and Agrion) and Hemiptera, with especial reference to the Embryonal integument" (blastoderm). With these two papers, the latter relating to the earliest changes in the eggs of Dragon-flies, and the former to the later stages in the life of the embryo, we have quite a complete account of the evolution of this remarkable family of insects. Dr. Brandt also gives the developmental history of certain Hemiptera (Corixa, Hydrometra, Lecanium and Aphis), and shows the remarkable identity in the embryology of these insects with the neuropterous insects mentioned above. A few other articles have appeared by Newport, Van Beneden, and others. We have already in the present volume of the NATURALIST, quoted from the abstract of Robins' paper on the "Development of Mites," quoted from the "Comptes Rendus" of the French Academy.

We would now notice the last work on the embryology of insects, that of Claparède, entitled "Studies on the Acarina," and published in Siebold and Kölliker's Journal of Scientific Zoology, during the present year. Claparède has observed in Atax Bonzi, which is a parasite on the gills of fresh-water mussels, that out of the originally laid egg (Pl. 8, fig. 3, embryo of Tyroglyphus siro, which closely resembles the earliest stages of the embryo of Atax; vt, yolk; md, mandibles; mx, maxillæ; p'-p''', legs. Fig. 4, front view of the same; R, beak; p, maxillæ), not a larva, but an egg-shaped form hatches, which he calls a "deutovum." (Pl. 8, fig. 1, bursting of the egg-shell into two halves, mo, on the day the deutovum, dm, hatches out; md, mandibles; mx, maxillæ; p', third pair of legs; lh, body cavity; sp, common beginning of the alimentary canal and nervous system; amb, hæmababa, amoeba-like bodies, which represent the blood corpuscles; there being no circulation of the blood, the movements of the hæmababa constitute a vicarious circulation. Fig. 2, the deutovum free from the first egg-shell; lettering same as in Fig. 1; oc, rudiments of the simple eyes; R, beak, hh', rudimentary stomach and liver). From this deutovum (which is not the "amnion" of insects), is developed a six-footed larva. This larva passes into an eight-footed form, the "second larva" (the "nymph" or pupa, of Dujardin and Robin) which transforms into the adult mite. The pupa differs from the adult in having longer feet, and four instead of ten genital clasping cups, the latter being the usual number in the adult.

The larvæ are elongated oval, with six long legs and four ocelli. They swarm over the gills of the mussel they are living on for a short time, and then bore into the substance of the gill to undergo their next transformation. Here the young mite increases in size, and becomes round. The tissues soften, those of the different organs not being so well marked as in the first larval stage. The limbs are short and much larger than before, the whole animal assuming an embryo-like appearance, and moving about like a rounded mass in its enclosure. Indeed is this process not (though Claparède does not say so) a histolysis of the former larval tissues, and the formation of a new body, as in the change of the six-footed insect beneath the larva skin where the pupa is formed? A new set of limbs grow out, this time there being four instead of three pairs of legs, while the old larval skin is still embraced within the membrane containing the second larval round mass. Soon the body is perfected, and the pupa, as we may properly call it, slips out of the larval membrane.

The "second larva" after some time undergoes another change; the limbs grow much shorter and are folded beneath the body, the animal being immovable, while the whole body assumes a broadly ovate form, and looks like an embryo just before hatching, but still lying within the egg. This may also be comparable with the formation of the adult fly within the puparium. (Compare Weismann's account of this process in Musca, in our "Guide to the Study of Insects," pp. 63, 64.) This period seems to be an exact repetition of the histolysis, and the formation of new tissues for the building up of a new body, which preceded the pupal stage, while the adult mite slips out of its pupal membrane, just as the pupal mite throws off its larval membrane, like an adult butterfly, or fly, emerging from its pupal membrane.

Thus the mites, at least several species, pass through a series of metamorphoses similar to those of such insects as have a complete metamorphosis (except that the Acarian pupa is active), while the absence of such metamorphosis in the spiders, is paralleled by the incomplete metamorphosis of the orthoptera and many neuroptera, which reach adult life by simple moultings of the skin.

In the genus Myobia there is not only a deutovum, besides the original egg, but also a tritovum-stage. The eggs of this mite are long, oval and conical at the posterior end. The embryo with the rudiments of limbs is represented by Fig. 5 of Plate 8. The little tubercles md and mx, represent the mandibles and maxillæ, while the three pairs of legs, p'p''p"", bud out from the middle of the body: lc represents the head-plate. The maxillæ and mandibles finally unite to form a beak (R, Fig. 6) and the three pairs of feet (p'p''p''') are folded along the median line of the body. The farther development of the embryo is now for a time arrested, and a peculiar tooth-like process (Fig. 7, d.) is developed. Claparède thinks that by means of this the anterior end of the egg-shell is cut off, and the embryo protrudes through, when (as in Fig. 7) it is seen to be surrounded by a new membrane, the deutovum (dt), equivalent to that of Atax. The