each side. An iridescence of great brilliancy is seen on the white surface of the abdomen and sides of the body. The head and upper portion of the body show a bluish, metallic sheen, and the tail, which is more or less flecked with brown or black, becomes in some specimens a bright red color. It would be difficult to exaggerate the beauty of coloring of these tadpoles, it exceeds in brilliancy and variety any species found in this locality.

As the legs become more fully developed, the coloring of the head and body tends from dark olive to a light, grayish-green. In the seventh week the body begins to lose its roundness, and the arms are seen to be moved under the skin, as if the tadpole were impatient to get them free. The head then appears disproportionately large. At this stage the tadpoles vary from gray to pea-green in color. They are found in the shallow water near the shore, where many fall prey to various aquatic birds. During the eighth week they appear to take little food; the arms are thrown out, the tail is gradually resorbed, the mouth developed, and the frogs leave the water. While a few specimens retain the color of gray up to this time, nearly all will be found of various shades of tender green on the upper surfaces, bordered with different tints of gray or salmon color. The abdomen is white. Green asserts itself much earlier in some specimens than others; but I have never seen a tadpole of this species develop into the frog that did not sooner or later become green. The markings on the back also vary in time of appearance; but the coloring of black on the head, body and limbs, the smooth shiny patch below the eyes, the granulated appearance of the skin, and the yellow coloring in the folds of the legs, usually appear in the order of their mention, and after the frogs have left the water.

Last season a small pond in an open pasture, about fifteen rods from a wood, furnished a good opportunity for observing their movements on leaving the water. From the 19th to the 24th of July, numbers of the young frogs, with tails in different stages of resorption, were found on the ground, weeds, and grasses about the pond, which by this time had become reduced by evaporation to a shallow pool. They represented a variety of shades of green; a few were gray, and occasionally one was scarcely to be separated in color from the mud on which it rested. I observed those on the ground frequently capture the small spiders which were numerous there. As soon as they left the water their object,

evidently, was to reach the wood. Apparently aware of their danger in this exposed journey, they drew attention to themselves, when approached, by continually springing out of harm's way; but after the shrubbery was reached they rarely made any attempt to escape when discovered, trusting wholly, like the mature frogs, to their disguise of coloring for safety. I found several of them on a small apple tree which was in the line of their journey. They were on the new growth which was overrun with Aphides, and the frogs had assumed a deep emerald-green, so like the leaf that it was difficult at first glance to distinguish them from it. After they reached the wood I could trace them no farther. I think it probable that some observers have mistaken H. versicolor at this for the adults of another species of Hyla.

age

My knowledge of the frogs from this stage till they reach maturity, is confined chiefly to those reared in a fernery. For the first three months they retained the green color, as a rule, with occasional changes to tints of brown and gray, matching the earth or branches to which they clung. After that time shades of gray became the rule and green the exception. The black markings on the head, body and limbs did not change excepting to vary in distinctness. Their food, which they never took unless alive, was Aphides at first, but soon flies formed their chief diet. During the day they commonly remained motionless, hidden behind the bark of the branches, with feet and hands, which are evidently extremely sensitive, compactly folded under the body, so that only their outer edges came in contact with the surface on which they were seated. Occasionally they would pat the disks against the sides of the body as if to moisten them. Their activity was reserved for the night, although rain accompanied by a south wind, caused them to move about uneasily. About the 1st of October they left the branches and ferns and nestled away in the damp earth and moss, where they remained through the winter, unless exposed to a temperature above 60°. They took no food from the first week in October till the 14th of the following May, when I gave them their liberty. They were then placed on an oak tree, where, after climbing till a suitable crevice or hiding place was found, they backed themselves into it and became to all appearance like a part of the bark of the tree.

ON SOME ENTOMOSTRACA OF LAKE MICHIGAN AND ADJACENT WATERS.

BY S. A. Forbes.

(Continued from July number.)

A THIRD Calanid deserves special mention as a species of Lim

nocalanus, a genus known hitherto only from Scandinavian lakes. It is readily distinguished, without dissection, from the other fresh-water Calanidæ, by the extraordinary length, size and prominence of the five or six terminal setæ of the first maxillipeds. The second maxillipeds are also very long. The legs are all bi-ramose, the inner ramus of the fifth pair resembling the same appendage of the other legs. This species, which may be Limnocalanus macrurus Sars, was first sent me by Mr. C. S. Fellows, of Chicago, about four years ago, a mutilated female having been obtained by him from the city water supply. The furca is as long as the entire abdomen. The rami are hairy, parallel, about seven times as long as wide, and provided with five subequal terminal setæ, and one some distance in front of the external angle. It has been collected thus far only in the south end of the lake. I have found it abundant in the city harbor, even in the polluted water near the mouth of the river, where it is associated especially with Diaptomus sicilis and the Cyclops next to be described.

The Calanidæ seem to have an unusual development in this country; and to facilitate their study and comparison, I have described further on all the species which I have hitherto clearly distinguished.

Smallest and most abundant of the Copepoda of the lake, is a minute Cyclops (C. thomasi, n. s., Pl. 1x Figs. 10, II and 16), only four hundredths of an inch in length (without setæ) and about eleven thousandths of an inch in width, slender and colorless, with remarkably long caudal stylets; and especially noticeable for the great difference in the length of the caudal setæ. The inner and outer ones are inconspicuous, while the outer of the two median setæ is longer than the furca, and the inner of these two is as long as the whole abdomen. This Cyclops was first received from Mr. B. W. Thomas, and I have since found it excessively abundant in the lake. I have not encountered it, however, in any other waters.

Its nearest European ally is apparently Cyclops bicuspidatus

Claus, but from this the description on another page will serve easily to distinguish it.

This is the only Cyclops which I have yet noticed in Lake Michigan, and is certainly far the most abundant species.

Of the many species of Cladocera occurring here, I have selected but three for especial comment. The first of these, Leptodora hyalina Lillj. (Pl. 1x Fig. 3), which occurs also in Europe, is a most interesting creature. When in its native element it is almost perfectly transparent, and consequently invisible-a true microscopic ghost-a fact associated by Professor Weissmann with its predaceous habit and feeble locomotive power. To the little Cyclops host it must indeed be a dreadful and mysterious enemy. Concealed by its transparency, it need not lurk in obscure hiding places, like grosser robbers, but can wing its way unnoticed among its prey.

The common Daphnids of the lake are, however, almost equally transparent, and as these are not at all carnivorous, we must either suppose that they have developed independently the same peculiarity for a directly opposite purpose-that of self-protection-or else we must conclude that there is something in the conditions of life here which tends to render the bodies of all entomostraca transparent.

A single mutilated specimen of Leptodora was dredged by Professor S. I. Smith in Lake Superior; it has been found in both ends of Lake Michigan, and I have also collected it in the Illinois river and the small lakes adjacent, and in a muddy pond in Northern Illinois only half a mile across and twenty feet in depth. A careful comparison of my specimens with the descriptions and figures of Lilljeborg and Weissmann, leaves no room for doubt that they belong to the European species.

This is likewise the case with the remarkable Holopedium gibberum Zaddach (Pl. Ix Figs. 12-15) found as yet only in Grand Traverse bay, where it occurred not rarely with Epischura, Diaptomus, Cyclops and Daphnia hyalina. In this animal the bivalvė shell has undergone a truly monstrous development, the brood cavity on the back being elevated to a height greater, when filled with young, than that of the remainder of the animal. On the other hand, the lateral valves of the shell are so shortened that they do not completely cover the branchial feet. For the protection of the creature and its young, and partly also, according to

P. E. Müller's supposition, to restore the balance of the body and enable it to float feet downwards, the shell secretes a layer or cloak of a gelatinous character and of an enormous thickness, relatively to the size of the animal. Through a slit in this mantle the antennæ and feet are thrust out; but otherwise the animal is completely buried in a lump of gelatine.

Bosmina (Pl. Ix Fig. 17) was less abundant in my collections than the other forms mentioned, but occurred very commonly in the stomachs of Mysis oculatus dredged from the deeper waters of the bay.

The commonest Cladocera in the lake are two forms of Daphnia, remarkable for their thinness and exquisite transparency. They are allied to galeata and pellucida of the old world, recently reduced by Adolph Lutz to varieties of Daphnia hyalina Leydig. Although our specimens do not agree strictly either with the descriptions or the figures of those varieties extant, their differences probably do not pass the limits of allowable variation in this excessively variable species. The head is keeled, convex in dorsal outline and either rounded (pellucida) or pointed (galeata) in front, the shell is compressed and reticulate, and terminates posteriorly in a long, straight, dentate spine.

An allied species, from the smaller Illinois lakes, where it is in autumn by far the most abundant entomostracan, resembles Daphnia cederströmi Schödler, but differs especially in the still more enormous development of the head. This is as high as the body and more than two-thirds as long, deeply concave on the upper border, the apex curving upwards far beyond the dorsal line of the body. The head is expanded inferiorly also to such a degree that the sensory hairs of the antennules fall much short of the tip of the rostrum. The shell is reticulate, and its spine long and straight, there is no macula nigra, and the caudal claws have a row of teeth at their base. For this curious form I propose the name of Daphnia retrocurva.

I have not found it in Lake Michigan, although in the smaller lakes it is mingled with both varieties of Daphnia hyalina. Even in the young, before they have left the brood cavity of the mother, the helmet is developed far beyond that of the adult of any of the latter species.

The female carries but one or two eggs, and the young sometimes attain a size more than half that of the body of the mother within the shell, before they leave her protection.