struggle for existence, and perhaps consider that the metamorphoses of the animal within the egg are due to a reflex action of the modes of life of the ancestors of the animal on the embryos of its descendants.

EXPLANATION OF PLATE 1.

Fig. 1. Colpocephalum lari Pack. 1a, antenna. The short line by the side gives the length of the insect.

Fig. 2. Lipeurus corvi Pack. 2a, antenna.

"3. Docophorus buteonis Pack. 3a, antenna.

"4. Lipeurus elongatus Pack. 4a, antenna. "5. Nirmus thoracicus Pack.

6. Lipeurus gracilis Pack.

"7. Docophorus hamatus Pack.

NOTES ON FRESH-WATER FISHES OF NEW JERSEY.

BY CHARLES C. ABBOTT, M. D.

THE character of the Delaware River, in the vicinity of Trenton, New Jersey, the head of navigation, is quite varied; the bed is stony, with scattered large rocks above the rapids, and sandy, with some vegetation below the falls; the current is swift to the rapids, but less so, being tide water, below them; these conditions, with that of the varied character of the tributaries at and near Trenton, make it an excellent point at which to examine the ichthyology of this river basin. This has been done partly by those who have received collections therefrom; but there is nothing in the publications of their studies giving any knowledge of the habits of these fish, but simply the fact of their presence in these

waters.

The ichthyic fauna is quite large, as some streams are cold and swift, that until lately harbored trout; and other streams, sluggish and thick, that are paradisiacal to the mudfish (Melanura), and the sucker (Hylomyzon).

In the present paper we propose not only to mention the results of the study of the habits of the species particularized, but to refer also to observations we have made, that apply to the fishes of these waters as a class, rather than to any single species. These observations we will give first, and then notice separately the more interesting species, in conclusion.

We would first call attention to alterations in circumscribed fauna. These changes are what have occurred lately in the small brooks, either emptying into the river directly, or tributary to the two large creeks, the Assunpink, and Crosswicks. We give only such instances as have occurred under our own notice. In the month of June, 1867, we fished the entire length of a never failing spring-brook, remarkable always for the number of specimens, if not of species. The fauna consisted, as usual, of chubs (Semotilus rhotheus and S. corporalis); dace (Argyreus atronasus), and minnows (Fundulus multifasciatus). The abundance of these species was relatively as named. During the first week of July following, a heavy, sudden fall of rain caused a considerable rise in the brook, and the extra bulk of water rushing over the narrow bed, altered the character of the brook so slightly, that it attracted no notice from those accustomed to seeing it daily. On the subsidence of the water, no cyprinoids, or in fact other fish, could be found, although we left hundreds in the stream. A week later we found a few roach (Stilbe Americana); they were never seen by us previously, in this stream, and still later, young mullet (Moxostoma oblongum). No chub have since been seen in this brook, which during the summer past (1869), was well tenanted with the species substituted in 1867, for them. During the last summer a few red-fins (Hypsilepis cornutus), and shiner (Hypsilepis Kentuckiensis), made their appearance. In a similar instance, happening in 1868, a familiar creek, teeming with cyprinoids, but with representatives of no other family, was found after a freshet to have lost a large number

of its species, and those remaining, represented by but few individuals; while percoids, heretofore wanting, appeared in the shape of Banded Sunfish (Bryttus chatodon), and Spotted-finned Sunfish (B. punctatus); also a few specimens of the Pirate (Aphrodederus Sayanus) were met with.

A third instance of alteration in the fauna, with no change in the bulk of water, occurred in the Shabbaconk Creek, a creek flowing into the Assunpink, which latter is dammed at its mouth, effectually preventing fish, leaving this creek, from returning to it. In this instance, the Aphrodederus Sayanus, which, for several seasons previous to 1867, had been abundant, suddenly disappeared. We have searched for them repeatedly since, but never have taken a single specimen. In the Assumpink Creek, where these "pirates" it would seem must have gone to, we have also carefully searched, but its extensive basin has not yet furnished a single specimen.

Such experiences of one familiar with these waters for fifteen years, explain why it is that different visitors in a few years examination of a stream or neighborhood, will in their reports differ considerably. One's own notes may be very inconsistent, on comparing those of any year with that of the preceding or following season. Even to the smaller cyprinoids, that are, we would suppose indisposed, if able, to migrate, we have applied the terms "abundant," "rare," "numerous," "scarce," at different times. More frequently these contradictory "remarks" were jotted down with reference to the occupants of small streams, but not altogether So. It is our custom now to look upon the contents of any one stream as but very imperfectly showing the fauna of that neighborhood, for two water-courses similar in all respects to the eye, may have no species common to each, although but two or three miles distant. In concluding what we have to say under this head of changes in faune-we would call attention to our experience in finding ourselves apparently or really in error. Frequently